Upload folder using huggingface_hub

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_512_llm/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_512_llm/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "k_proj",
+    "v_proj",
+    "gate_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_512_llm/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a87db262e98cd4b8d78beb4f9e263fb293df05635b3ac4e3c51dd666bea98e88
+size 1384057050

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_512_llm/config.json

@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_512_llm/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_512_llm/non_lora_trainables.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
+size 912

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_512_llm/trainer_state.json

@@ -0,0 +1,3530 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.85112111944993,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.4169,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.349335609541988,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.4395,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8901132050817202,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.5624,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7730611136792902,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.5882,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.58151584131742,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.4787,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.655609159821099,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8102,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5330182754444979,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.4769,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4708190466525344,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.5067,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1741416125033164,
+      "learning_rate": 6e-06,
+      "loss": 0.4245,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4884515435695005,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.4758,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9273383382028549,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.3257,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2746507856751552,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.4055,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.014594518697626,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.5426,
+      "step": 13
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8586569224695126,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.4112,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1219716136449773,
+      "learning_rate": 1e-05,
+      "loss": 0.2521,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1573810523787038,
+      "learning_rate": 9.999895105006995e-06,
+      "loss": 0.2143,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7453793880408,
+      "learning_rate": 9.99958042442916e-06,
+      "loss": 0.292,
+      "step": 17
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.132964534166692,
+      "learning_rate": 9.999055971469864e-06,
+      "loss": 0.2195,
+      "step": 18
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.129476208393603,
+      "learning_rate": 9.998321768134101e-06,
+      "loss": 0.2723,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9509982409945354,
+      "learning_rate": 9.997377845227577e-06,
+      "loss": 0.2238,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2039057738144747,
+      "learning_rate": 9.9962242423554e-06,
+      "loss": 0.2227,
+      "step": 21
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8827305723192147,
+      "learning_rate": 9.99486100792044e-06,
+      "loss": 0.1864,
+      "step": 22
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9315903551587646,
+      "learning_rate": 9.993288199121283e-06,
+      "loss": 0.1668,
+      "step": 23
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6851374572709672,
+      "learning_rate": 9.991505881949837e-06,
+      "loss": 0.1307,
+      "step": 24
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6909485913620177,
+      "learning_rate": 9.98951413118856e-06,
+      "loss": 0.16,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7406648593699517,
+      "learning_rate": 9.987313030407325e-06,
+      "loss": 0.1837,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7238996498107553,
+      "learning_rate": 9.984902671959911e-06,
+      "loss": 0.1254,
+      "step": 27
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.036829853615859,
+      "learning_rate": 9.982283156980133e-06,
+      "loss": 0.1257,
+      "step": 28
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46928883832662704,
+      "learning_rate": 9.979454595377594e-06,
+      "loss": 0.1402,
+      "step": 29
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7995129150340999,
+      "learning_rate": 9.97641710583307e-06,
+      "loss": 0.133,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7158087581331645,
+      "learning_rate": 9.973170815793543e-06,
+      "loss": 0.1721,
+      "step": 31
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1000323699939043,
+      "learning_rate": 9.969715861466839e-06,
+      "loss": 0.2289,
+      "step": 32
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7540364135170671,
+      "learning_rate": 9.966052387815923e-06,
+      "loss": 0.1731,
+      "step": 33
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9083373383705893,
+      "learning_rate": 9.962180548552812e-06,
+      "loss": 0.1548,
+      "step": 34
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7013209781158779,
+      "learning_rate": 9.958100506132127e-06,
+      "loss": 0.1093,
+      "step": 35
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41412415464083024,
+      "learning_rate": 9.953812431744274e-06,
+      "loss": 0.1133,
+      "step": 36
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9425192972976834,
+      "learning_rate": 9.94931650530827e-06,
+      "loss": 0.1646,
+      "step": 37
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.675296221063655,
+      "learning_rate": 9.944612915464183e-06,
+      "loss": 0.1718,
+      "step": 38
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8388610957624392,
+      "learning_rate": 9.93970185956522e-06,
+      "loss": 0.1141,
+      "step": 39
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5596739992412059,
+      "learning_rate": 9.934583543669454e-06,
+      "loss": 0.1183,
+      "step": 40
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7061961098399181,
+      "learning_rate": 9.929258182531167e-06,
+      "loss": 0.1445,
+      "step": 41
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8146238570997492,
+      "learning_rate": 9.923725999591846e-06,
+      "loss": 0.13,
+      "step": 42
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6982418193712115,
+      "learning_rate": 9.917987226970811e-06,
+      "loss": 0.1887,
+      "step": 43
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45737000411433804,
+      "learning_rate": 9.912042105455462e-06,
+      "loss": 0.1061,
+      "step": 44
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49781720686446135,
+      "learning_rate": 9.905890884491196e-06,
+      "loss": 0.133,
+      "step": 45
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5911717837670529,
+      "learning_rate": 9.899533822170922e-06,
+      "loss": 0.1108,
+      "step": 46
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.762707651269511,
+      "learning_rate": 9.892971185224244e-06,
+      "loss": 0.1585,
+      "step": 47
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6655950445860507,
+      "learning_rate": 9.886203249006265e-06,
+      "loss": 0.1677,
+      "step": 48
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3521712122009652,
+      "learning_rate": 9.879230297486034e-06,
+      "loss": 0.0948,
+      "step": 49
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6177305571716366,
+      "learning_rate": 9.872052623234632e-06,
+      "loss": 0.1569,
+      "step": 50
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6707310128698482,
+      "learning_rate": 9.864670527412891e-06,
+      "loss": 0.1821,
+      "step": 51
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7363677911996476,
+      "learning_rate": 9.857084319758772e-06,
+      "loss": 0.1538,
+      "step": 52
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7240548069315486,
+      "learning_rate": 9.849294318574353e-06,
+      "loss": 0.1462,
+      "step": 53
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.516881727794041,
+      "learning_rate": 9.841300850712479e-06,
+      "loss": 0.1053,
+      "step": 54
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46953360641861996,
+      "learning_rate": 9.833104251563058e-06,
+      "loss": 0.1348,
+      "step": 55
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.593834579076815,
+      "learning_rate": 9.824704865038967e-06,
+      "loss": 0.1406,
+      "step": 56
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7140035131215059,
+      "learning_rate": 9.816103043561648e-06,
+      "loss": 0.1422,
+      "step": 57
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8616675617901599,
+      "learning_rate": 9.807299148046301e-06,
+      "loss": 0.1028,
+      "step": 58
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7572015385538562,
+      "learning_rate": 9.798293547886748e-06,
+      "loss": 0.1306,
+      "step": 59
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.505293583960689,
+      "learning_rate": 9.789086620939936e-06,
+      "loss": 0.1091,
+      "step": 60
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4393634667334124,
+      "learning_rate": 9.779678753510082e-06,
+      "loss": 0.1162,
+      "step": 61
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0255563164443287,
+      "learning_rate": 9.770070340332457e-06,
+      "loss": 0.134,
+      "step": 62
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6366914364146933,
+      "learning_rate": 9.76026178455684e-06,
+      "loss": 0.0967,
+      "step": 63
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8055640228185079,
+      "learning_rate": 9.75025349773058e-06,
+      "loss": 0.1686,
+      "step": 64
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6124730219998783,
+      "learning_rate": 9.740045899781353e-06,
+      "loss": 0.1231,
+      "step": 65
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3953777914303892,
+      "learning_rate": 9.729639418999524e-06,
+      "loss": 0.1056,
+      "step": 66
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5690459588115772,
+      "learning_rate": 9.719034492020183e-06,
+      "loss": 0.0993,
+      "step": 67
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.663681633150344,
+      "learning_rate": 9.708231563804828e-06,
+      "loss": 0.1537,
+      "step": 68
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5078103691237422,
+      "learning_rate": 9.697231087622691e-06,
+      "loss": 0.0793,
+      "step": 69
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5544023432875107,
+      "learning_rate": 9.68603352503172e-06,
+      "loss": 0.1162,
+      "step": 70
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5887817014329297,
+      "learning_rate": 9.674639345859213e-06,
+      "loss": 0.157,
+      "step": 71
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5573720967228267,
+      "learning_rate": 9.663049028182112e-06,
+      "loss": 0.1363,
+      "step": 72
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5293879592609317,
+      "learning_rate": 9.651263058306932e-06,
+      "loss": 0.1177,
+      "step": 73
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6349288469559691,
+      "learning_rate": 9.639281930749363e-06,
+      "loss": 0.1289,
+      "step": 74
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5389342135849522,
+      "learning_rate": 9.627106148213521e-06,
+      "loss": 0.1335,
+      "step": 75
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6030463446293332,
+      "learning_rate": 9.61473622157086e-06,
+      "loss": 0.1357,
+      "step": 76
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.845013424171041,
+      "learning_rate": 9.602172669838721e-06,
+      "loss": 0.1331,
+      "step": 77
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6778027972809342,
+      "learning_rate": 9.589416020158577e-06,
+      "loss": 0.0741,
+      "step": 78
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3897778551366759,
+      "learning_rate": 9.5764668077739e-06,
+      "loss": 0.093,
+      "step": 79
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5717939081881331,
+      "learning_rate": 9.563325576007702e-06,
+      "loss": 0.0946,
+      "step": 80
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39021049181588896,
+      "learning_rate": 9.549992876239753e-06,
+      "loss": 0.0824,
+      "step": 81
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.873387309542934,
+      "learning_rate": 9.536469267883432e-06,
+      "loss": 0.1242,
+      "step": 82
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5796663877543419,
+      "learning_rate": 9.52275531836226e-06,
+      "loss": 0.1299,
+      "step": 83
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.54252944507829,
+      "learning_rate": 9.508851603086094e-06,
+      "loss": 0.1249,
+      "step": 84
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.372797578486964,
+      "learning_rate": 9.494758705426978e-06,
+      "loss": 0.1441,
+      "step": 85
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5636216727890436,
+      "learning_rate": 9.480477216694674e-06,
+      "loss": 0.1091,
+      "step": 86
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8752904523458626,
+      "learning_rate": 9.466007736111846e-06,
+      "loss": 0.1807,
+      "step": 87
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6014787546432877,
+      "learning_rate": 9.451350870788922e-06,
+      "loss": 0.1026,
+      "step": 88
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8350725506944692,
+      "learning_rate": 9.436507235698613e-06,
+      "loss": 0.1641,
+      "step": 89
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0005403108385666,
+      "learning_rate": 9.421477453650118e-06,
+      "loss": 0.1163,
+      "step": 90
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6716285942507424,
+      "learning_rate": 9.406262155262995e-06,
+      "loss": 0.138,
+      "step": 91
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6438179692572337,
+      "learning_rate": 9.390861978940687e-06,
+      "loss": 0.1249,
+      "step": 92
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6596989843894778,
+      "learning_rate": 9.37527757084375e-06,
+      "loss": 0.1037,
+      "step": 93
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5493756528846595,
+      "learning_rate": 9.359509584862735e-06,
+      "loss": 0.1203,
+      "step": 94
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6279119986783067,
+      "learning_rate": 9.343558682590757e-06,
+      "loss": 0.1399,
+      "step": 95
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.574686461780649,
+      "learning_rate": 9.327425533295725e-06,
+      "loss": 0.1416,
+      "step": 96
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5928939130040762,
+      "learning_rate": 9.31111081389227e-06,
+      "loss": 0.1003,
+      "step": 97
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8303152054811029,
+      "learning_rate": 9.29461520891335e-06,
+      "loss": 0.1095,
+      "step": 98
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.744634256418121,
+      "learning_rate": 9.277939410481507e-06,
+      "loss": 0.1125,
+      "step": 99
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5926044414182606,
+      "learning_rate": 9.261084118279846e-06,
+      "loss": 0.1198,
+      "step": 100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.589345092224048,
+      "learning_rate": 9.244050039522673e-06,
+      "loss": 0.1258,
+      "step": 101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4495956768162774,
+      "learning_rate": 9.226837888925813e-06,
+      "loss": 0.1405,
+      "step": 102
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.64920158414816,
+      "learning_rate": 9.209448388676636e-06,
+      "loss": 0.0981,
+      "step": 103
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44865801609533834,
+      "learning_rate": 9.191882268403743e-06,
+      "loss": 0.093,
+      "step": 104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5304956297983962,
+      "learning_rate": 9.174140265146356e-06,
+      "loss": 0.134,
+      "step": 105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7762471405689545,
+      "learning_rate": 9.156223123323405e-06,
+      "loss": 0.1152,
+      "step": 106
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5898727313393679,
+      "learning_rate": 9.13813159470227e-06,
+      "loss": 0.1283,
+      "step": 107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4616139402394573,
+      "learning_rate": 9.119866438367263e-06,
+      "loss": 0.0955,
+      "step": 108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4975289379033519,
+      "learning_rate": 9.101428420687759e-06,
+      "loss": 0.0832,
+      "step": 109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4265320746554224,
+      "learning_rate": 9.082818315286054e-06,
+      "loss": 0.081,
+      "step": 110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5862894990592699,
+      "learning_rate": 9.0640369030049e-06,
+      "loss": 0.1404,
+      "step": 111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5585869181490479,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.1175,
+      "step": 112
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9547603316749469,
+      "learning_rate": 9.025963317080641e-06,
+      "loss": 0.1593,
+      "step": 113
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4493417664599385,
+      "learning_rate": 9.006672740928952e-06,
+      "loss": 0.1034,
+      "step": 114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4508339845547332,
+      "learning_rate": 8.987214052813605e-06,
+      "loss": 0.0869,
+      "step": 115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.47524415745527615,
+      "learning_rate": 8.967588069182184e-06,
+      "loss": 0.1193,
+      "step": 116
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5773831975479189,
+      "learning_rate": 8.947795613501658e-06,
+      "loss": 0.1061,
+      "step": 117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.697932596256695,
+      "learning_rate": 8.927837516223824e-06,
+      "loss": 0.0871,
+      "step": 118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1421056116230779,
+      "learning_rate": 8.907714614750473e-06,
+      "loss": 0.1202,
+      "step": 119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2641030779009608,
+      "learning_rate": 8.887427753398249e-06,
+      "loss": 0.1328,
+      "step": 120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6361326163766521,
+      "learning_rate": 8.866977783363219e-06,
+      "loss": 0.1359,
+      "step": 121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5093836960908875,
+      "learning_rate": 8.846365562685178e-06,
+      "loss": 0.1131,
+      "step": 122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6190698249115906,
+      "learning_rate": 8.825591956211614e-06,
+      "loss": 0.1556,
+      "step": 123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7541491658108361,
+      "learning_rate": 8.804657835561456e-06,
+      "loss": 0.1323,
+      "step": 124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.647720482115624,
+      "learning_rate": 8.783564079088478e-06,
+      "loss": 0.0978,
+      "step": 125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.669906338449493,
+      "learning_rate": 8.762311571844453e-06,
+      "loss": 0.1404,
+      "step": 126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9522809597720677,
+      "learning_rate": 8.74090120554202e-06,
+      "loss": 0.0856,
+      "step": 127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7776506061986043,
+      "learning_rate": 8.719333878517274e-06,
+      "loss": 0.1236,
+      "step": 128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4248528641074464,
+      "learning_rate": 8.697610495692055e-06,
+      "loss": 0.0934,
+      "step": 129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6747066450629217,
+      "learning_rate": 8.675731968536004e-06,
+      "loss": 0.1082,
+      "step": 130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5041768519333474,
+      "learning_rate": 8.653699215028298e-06,
+      "loss": 0.107,
+      "step": 131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8104039298114765,
+      "learning_rate": 8.63151315961915e-06,
+      "loss": 0.1251,
+      "step": 132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5705359883870549,
+      "learning_rate": 8.609174733191012e-06,
+      "loss": 0.1683,
+      "step": 133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5967376334947457,
+      "learning_rate": 8.586684873019513e-06,
+      "loss": 0.1194,
+      "step": 134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6584709702471544,
+      "learning_rate": 8.564044522734147e-06,
+      "loss": 0.1303,
+      "step": 135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4553899784083641,
+      "learning_rate": 8.541254632278667e-06,
+      "loss": 0.1015,
+      "step": 136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4542733986833433,
+      "learning_rate": 8.518316157871232e-06,
+      "loss": 0.1199,
+      "step": 137
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.36334886125035926,
+      "learning_rate": 8.495230061964289e-06,
+      "loss": 0.1212,
+      "step": 138
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4907989038266779,
+      "learning_rate": 8.471997313204183e-06,
+      "loss": 0.1197,
+      "step": 139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.672024265497057,
+      "learning_rate": 8.448618886390523e-06,
+      "loss": 0.1297,
+      "step": 140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39193548778136306,
+      "learning_rate": 8.425095762435274e-06,
+      "loss": 0.0973,
+      "step": 141
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4692220452901667,
+      "learning_rate": 8.401428928321607e-06,
+      "loss": 0.1107,
+      "step": 142
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6001012390203628,
+      "learning_rate": 8.377619377062483e-06,
+      "loss": 0.1499,
+      "step": 143
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41281023311178233,
+      "learning_rate": 8.353668107658984e-06,
+      "loss": 0.1132,
+      "step": 144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4997548913764334,
+      "learning_rate": 8.329576125058406e-06,
+      "loss": 0.1393,
+      "step": 145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9491246508670012,
+      "learning_rate": 8.305344440112089e-06,
+      "loss": 0.093,
+      "step": 146
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6368389591245665,
+      "learning_rate": 8.280974069532999e-06,
+      "loss": 0.1221,
+      "step": 147
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8801854448376024,
+      "learning_rate": 8.256466035853077e-06,
+      "loss": 0.1376,
+      "step": 148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46974334183739896,
+      "learning_rate": 8.231821367380335e-06,
+      "loss": 0.0883,
+      "step": 149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4563432085844253,
+      "learning_rate": 8.207041098155701e-06,
+      "loss": 0.0876,
+      "step": 150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5143474212167232,
+      "learning_rate": 8.182126267909642e-06,
+      "loss": 0.0919,
+      "step": 151
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6064788787568761,
+      "learning_rate": 8.157077922018537e-06,
+      "loss": 0.1254,
+      "step": 152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8717312999650924,
+      "learning_rate": 8.13189711146081e-06,
+      "loss": 0.1577,
+      "step": 153
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7624938303459425,
+      "learning_rate": 8.106584892772844e-06,
+      "loss": 0.1208,
+      "step": 154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49154776786386456,
+      "learning_rate": 8.081142328004638e-06,
+      "loss": 0.1392,
+      "step": 155
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6282719321510984,
+      "learning_rate": 8.055570484675252e-06,
+      "loss": 0.1262,
+      "step": 156
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7673100285994615,
+      "learning_rate": 8.029870435728018e-06,
+      "loss": 0.1251,
+      "step": 157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6854992028342847,
+      "learning_rate": 8.004043259485519e-06,
+      "loss": 0.1399,
+      "step": 158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8294681941356243,
+      "learning_rate": 7.978090039604342e-06,
+      "loss": 0.1223,
+      "step": 159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5036842237563715,
+      "learning_rate": 7.952011865029614e-06,
+      "loss": 0.1042,
+      "step": 160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6473097444078603,
+      "learning_rate": 7.925809829949312e-06,
+      "loss": 0.1373,
+      "step": 161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5425935965563373,
+      "learning_rate": 7.89948503374835e-06,
+      "loss": 0.1054,
+      "step": 162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4622580056148776,
+      "learning_rate": 7.873038580962453e-06,
+      "loss": 0.1299,
+      "step": 163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4647795187819019,
+      "learning_rate": 7.846471581231814e-06,
+      "loss": 0.1011,
+      "step": 164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32064580771685813,
+      "learning_rate": 7.819785149254534e-06,
+      "loss": 0.0859,
+      "step": 165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37761395521574226,
+      "learning_rate": 7.792980404739849e-06,
+      "loss": 0.0951,
+      "step": 166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5879483997153534,
+      "learning_rate": 7.766058472361154e-06,
+      "loss": 0.1094,
+      "step": 167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8113614809532544,
+      "learning_rate": 7.739020481708816e-06,
+      "loss": 0.1236,
+      "step": 168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48352070058327146,
+      "learning_rate": 7.711867567242769e-06,
+      "loss": 0.1238,
+      "step": 169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.307579220599265,
+      "learning_rate": 7.68460086824492e-06,
+      "loss": 0.0789,
+      "step": 170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3052462933757519,
+      "learning_rate": 7.657221528771352e-06,
+      "loss": 0.0826,
+      "step": 171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4437060013882075,
+      "learning_rate": 7.629730697604314e-06,
+      "loss": 0.0956,
+      "step": 172
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5722033809772273,
+      "learning_rate": 7.602129528204023e-06,
+      "loss": 0.1377,
+      "step": 173
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7758402546865415,
+      "learning_rate": 7.574419178660269e-06,
+      "loss": 0.131,
+      "step": 174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9501235850355042,
+      "learning_rate": 7.546600811643816e-06,
+      "loss": 0.1662,
+      "step": 175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.36934033674544414,
+      "learning_rate": 7.5186755943576324e-06,
+      "loss": 0.0929,
+      "step": 176
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5281959851316647,
+      "learning_rate": 7.490644698487909e-06,
+      "loss": 0.1002,
+      "step": 177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34245725127334536,
+      "learning_rate": 7.462509300154892e-06,
+      "loss": 0.0887,
+      "step": 178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4866946829432723,
+      "learning_rate": 7.434270579863549e-06,
+      "loss": 0.063,
+      "step": 179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6814076859633514,
+      "learning_rate": 7.405929722454026e-06,
+      "loss": 0.0961,
+      "step": 180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7302450675973802,
+      "learning_rate": 7.3774879170519386e-06,
+      "loss": 0.172,
+      "step": 181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45339459954091,
+      "learning_rate": 7.348946357018479e-06,
+      "loss": 0.0925,
+      "step": 182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4956461896960731,
+      "learning_rate": 7.320306239900343e-06,
+      "loss": 0.0903,
+      "step": 183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6993675238341859,
+      "learning_rate": 7.291568767379484e-06,
+      "loss": 0.1453,
+      "step": 184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5283639812348182,
+      "learning_rate": 7.262735145222696e-06,
+      "loss": 0.1097,
+      "step": 185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7313507146714898,
+      "learning_rate": 7.233806583231012e-06,
+      "loss": 0.1371,
+      "step": 186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4652143703796043,
+      "learning_rate": 7.204784295188959e-06,
+      "loss": 0.0934,
+      "step": 187
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8693530781527562,
+      "learning_rate": 7.1756694988136165e-06,
+      "loss": 0.1106,
+      "step": 188
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5993104635573092,
+      "learning_rate": 7.14646341570353e-06,
+      "loss": 0.1078,
+      "step": 189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6311176939281898,
+      "learning_rate": 7.117167271287453e-06,
+      "loss": 0.1482,
+      "step": 190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46632768555288145,
+      "learning_rate": 7.0877822947729265e-06,
+      "loss": 0.1217,
+      "step": 191
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.380353772120599,
+      "learning_rate": 7.05830971909472e-06,
+      "loss": 0.1093,
+      "step": 192
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2314995066821384,
+      "learning_rate": 7.028750780863078e-06,
+      "loss": 0.1374,
+      "step": 193
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3359839639423677,
+      "learning_rate": 6.999106720311846e-06,
+      "loss": 0.0736,
+      "step": 194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30797035180919163,
+      "learning_rate": 6.969378781246436e-06,
+      "loss": 0.059,
+      "step": 195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3902264668766994,
+      "learning_rate": 6.939568210991633e-06,
+      "loss": 0.0803,
+      "step": 196
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4924002549898853,
+      "learning_rate": 6.9096762603392595e-06,
+      "loss": 0.1288,
+      "step": 197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45893113202500546,
+      "learning_rate": 6.8797041834956955e-06,
+      "loss": 0.0421,
+      "step": 198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3626386184662733,
+      "learning_rate": 6.849653238029261e-06,
+      "loss": 0.0868,
+      "step": 199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6171909067196144,
+      "learning_rate": 6.819524684817439e-06,
+      "loss": 0.1303,
+      "step": 200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49550909029618534,
+      "learning_rate": 6.78931978799398e-06,
+      "loss": 0.1094,
+      "step": 201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44111730247487485,
+      "learning_rate": 6.7590398148958625e-06,
+      "loss": 0.1082,
+      "step": 202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3945966419537577,
+      "learning_rate": 6.728686036010115e-06,
+      "loss": 0.1169,
+      "step": 203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3837774645373548,
+      "learning_rate": 6.698259724920503e-06,
+      "loss": 0.076,
+      "step": 204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.451749688800484,
+      "learning_rate": 6.667762158254104e-06,
+      "loss": 0.1018,
+      "step": 205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3504727281000213,
+      "learning_rate": 6.637194615627733e-06,
+      "loss": 0.0852,
+      "step": 206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40554396216881844,
+      "learning_rate": 6.6065583795942625e-06,
+      "loss": 0.0954,
+      "step": 207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5522740050815659,
+      "learning_rate": 6.5758547355887944e-06,
+      "loss": 0.0895,
+      "step": 208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6968752850834635,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.1395,
+      "step": 209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5882665903589881,
+      "learning_rate": 6.514250379489754e-06,
+      "loss": 0.1246,
+      "step": 210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5672951512318772,
+      "learning_rate": 6.483352252191585e-06,
+      "loss": 0.1323,
+      "step": 211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5661159807973724,
+      "learning_rate": 6.452391886403767e-06,
+      "loss": 0.1178,
+      "step": 212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34353828911630974,
+      "learning_rate": 6.421370581161244e-06,
+      "loss": 0.1009,
+      "step": 213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5407518626154225,
+      "learning_rate": 6.390289638055851e-06,
+      "loss": 0.137,
+      "step": 214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.372311277153281,
+      "learning_rate": 6.3591503611817155e-06,
+      "loss": 0.0926,
+      "step": 215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5542289788795887,
+      "learning_rate": 6.3279540570805265e-06,
+      "loss": 0.1081,
+      "step": 216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4976922442704691,
+      "learning_rate": 6.296702034686726e-06,
+      "loss": 0.0953,
+      "step": 217
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40534247934992484,
+      "learning_rate": 6.265395605272581e-06,
+      "loss": 0.1395,
+      "step": 218
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4097713453292996,
+      "learning_rate": 6.234036082393171e-06,
+      "loss": 0.1139,
+      "step": 219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4799177841662269,
+      "learning_rate": 6.202624781831269e-06,
+      "loss": 0.0886,
+      "step": 220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5460833433496859,
+      "learning_rate": 6.171163021542134e-06,
+      "loss": 0.1203,
+      "step": 221
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5522116184374418,
+      "learning_rate": 6.139652121598219e-06,
+      "loss": 0.1101,
+      "step": 222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6140517613951189,
+      "learning_rate": 6.108093404133772e-06,
+      "loss": 0.1547,
+      "step": 223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45193523540509095,
+      "learning_rate": 6.076488193289375e-06,
+      "loss": 0.0951,
+      "step": 224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5965337969169188,
+      "learning_rate": 6.044837815156377e-06,
+      "loss": 0.1311,
+      "step": 225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8953228703641515,
+      "learning_rate": 6.013143597721252e-06,
+      "loss": 0.1274,
+      "step": 226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.550454631947633,
+      "learning_rate": 5.981406870809889e-06,
+      "loss": 0.1343,
+      "step": 227
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5338311597923737,
+      "learning_rate": 5.949628966031785e-06,
+      "loss": 0.11,
+      "step": 228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5420821295239983,
+      "learning_rate": 5.9178112167241805e-06,
+      "loss": 0.1356,
+      "step": 229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5236703385396021,
+      "learning_rate": 5.885954957896115e-06,
+      "loss": 0.1505,
+      "step": 230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4649375414577938,
+      "learning_rate": 5.854061526172402e-06,
+      "loss": 0.1118,
+      "step": 231
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9506605375806841,
+      "learning_rate": 5.822132259737565e-06,
+      "loss": 0.1487,
+      "step": 232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5025233378993809,
+      "learning_rate": 5.7901684982796716e-06,
+      "loss": 0.101,
+      "step": 233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45769338661823605,
+      "learning_rate": 5.75817158293414e-06,
+      "loss": 0.1268,
+      "step": 234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4738441021558865,
+      "learning_rate": 5.726142856227453e-06,
+      "loss": 0.0938,
+      "step": 235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41054890800736454,
+      "learning_rate": 5.694083662020835e-06,
+      "loss": 0.075,
+      "step": 236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4638293745171702,
+      "learning_rate": 5.661995345453867e-06,
+      "loss": 0.0983,
+      "step": 237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5020144058900003,
+      "learning_rate": 5.629879252888046e-06,
+      "loss": 0.1067,
+      "step": 238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6010730833165813,
+      "learning_rate": 5.597736731850295e-06,
+      "loss": 0.1303,
+      "step": 239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3522016094734147,
+      "learning_rate": 5.5655691309764225e-06,
+      "loss": 0.0908,
+      "step": 240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5919017521737345,
+      "learning_rate": 5.533377799954532e-06,
+      "loss": 0.1136,
+      "step": 241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5173116471794423,
+      "learning_rate": 5.501164089468406e-06,
+      "loss": 0.1322,
+      "step": 242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5388813789870869,
+      "learning_rate": 5.4689293511408155e-06,
+      "loss": 0.1246,
+      "step": 243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.45654642602275797,
+      "learning_rate": 5.43667493747682e-06,
+      "loss": 0.1033,
+      "step": 244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6132951284982724,
+      "learning_rate": 5.404402201807022e-06,
+      "loss": 0.1363,
+      "step": 245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4102608347812018,
+      "learning_rate": 5.372112498230771e-06,
+      "loss": 0.1123,
+      "step": 246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5181050047880049,
+      "learning_rate": 5.339807181559359e-06,
+      "loss": 0.1223,
+      "step": 247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5210515357795024,
+      "learning_rate": 5.307487607259175e-06,
+      "loss": 0.0874,
+      "step": 248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44620564412110364,
+      "learning_rate": 5.275155131394825e-06,
+      "loss": 0.078,
+      "step": 249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39212059919163095,
+      "learning_rate": 5.242811110572243e-06,
+      "loss": 0.0882,
+      "step": 250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.46051329484229414,
+      "learning_rate": 5.210456901881761e-06,
+      "loss": 0.1141,
+      "step": 251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43488813941628773,
+      "learning_rate": 5.1780938628411795e-06,
+      "loss": 0.0969,
+      "step": 252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3930932390650295,
+      "learning_rate": 5.145723351338799e-06,
+      "loss": 0.0919,
+      "step": 253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4133656343207384,
+      "learning_rate": 5.11334672557645e-06,
+      "loss": 0.1005,
+      "step": 254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44362256572085595,
+      "learning_rate": 5.080965344012509e-06,
+      "loss": 0.1159,
+      "step": 255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5489339948638737,
+      "learning_rate": 5.048580565304887e-06,
+      "loss": 0.1074,
+      "step": 256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4168118459374411,
+      "learning_rate": 5.016193748254045e-06,
+      "loss": 0.0882,
+      "step": 257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4391250559840595,
+      "learning_rate": 4.983806251745958e-06,
+      "loss": 0.1232,
+      "step": 258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9289464765767339,
+      "learning_rate": 4.951419434695115e-06,
+      "loss": 0.1453,
+      "step": 259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8668939729555174,
+      "learning_rate": 4.919034655987493e-06,
+      "loss": 0.1185,
+      "step": 260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.631274789168367,
+      "learning_rate": 4.886653274423551e-06,
+      "loss": 0.1354,
+      "step": 261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49360834913385504,
+      "learning_rate": 4.8542766486612035e-06,
+      "loss": 0.1099,
+      "step": 262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49736414112839294,
+      "learning_rate": 4.821906137158822e-06,
+      "loss": 0.1471,
+      "step": 263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3905951742394158,
+      "learning_rate": 4.7895430981182415e-06,
+      "loss": 0.083,
+      "step": 264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5205011969624558,
+      "learning_rate": 4.757188889427761e-06,
+      "loss": 0.1161,
+      "step": 265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.32771729503118857,
+      "learning_rate": 4.724844868605176e-06,
+      "loss": 0.0911,
+      "step": 266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33594696910541266,
+      "learning_rate": 4.6925123927408265e-06,
+      "loss": 0.0674,
+      "step": 267
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3438942174508842,
+      "learning_rate": 4.660192818440642e-06,
+      "loss": 0.0967,
+      "step": 268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43480873166148254,
+      "learning_rate": 4.627887501769231e-06,
+      "loss": 0.1269,
+      "step": 269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4297920266270292,
+      "learning_rate": 4.59559779819298e-06,
+      "loss": 0.0905,
+      "step": 270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4486852472418763,
+      "learning_rate": 4.5633250625231806e-06,
+      "loss": 0.128,
+      "step": 271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5822438467001619,
+      "learning_rate": 4.531070648859186e-06,
+      "loss": 0.1121,
+      "step": 272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4050734687784252,
+      "learning_rate": 4.498835910531595e-06,
+      "loss": 0.1029,
+      "step": 273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5737463360637343,
+      "learning_rate": 4.4666222000454685e-06,
+      "loss": 0.1139,
+      "step": 274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.823054947023667,
+      "learning_rate": 4.434430869023579e-06,
+      "loss": 0.1402,
+      "step": 275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4445981984272266,
+      "learning_rate": 4.402263268149707e-06,
+      "loss": 0.1091,
+      "step": 276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39441608036981185,
+      "learning_rate": 4.370120747111956e-06,
+      "loss": 0.104,
+      "step": 277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36168565396456787,
+      "learning_rate": 4.338004654546136e-06,
+      "loss": 0.074,
+      "step": 278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4388764114995804,
+      "learning_rate": 4.3059163379791676e-06,
+      "loss": 0.1276,
+      "step": 279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6445229540027669,
+      "learning_rate": 4.27385714377255e-06,
+      "loss": 0.1535,
+      "step": 280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.396872953907858,
+      "learning_rate": 4.24182841706586e-06,
+      "loss": 0.0652,
+      "step": 281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37145594860347253,
+      "learning_rate": 4.209831501720328e-06,
+      "loss": 0.0852,
+      "step": 282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.398644318684909,
+      "learning_rate": 4.177867740262437e-06,
+      "loss": 0.1428,
+      "step": 283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5249283308083765,
+      "learning_rate": 4.145938473827598e-06,
+      "loss": 0.0999,
+      "step": 284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6596209256901557,
+      "learning_rate": 4.1140450421038865e-06,
+      "loss": 0.1258,
+      "step": 285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.45806459464760085,
+      "learning_rate": 4.08218878327582e-06,
+      "loss": 0.0928,
+      "step": 286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5142127257187877,
+      "learning_rate": 4.050371033968216e-06,
+      "loss": 0.0948,
+      "step": 287
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5909032090575379,
+      "learning_rate": 4.018593129190113e-06,
+      "loss": 0.1291,
+      "step": 288
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4193073457818887,
+      "learning_rate": 3.98685640227875e-06,
+      "loss": 0.093,
+      "step": 289
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5717516521391793,
+      "learning_rate": 3.955162184843625e-06,
+      "loss": 0.0896,
+      "step": 290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5015285770752144,
+      "learning_rate": 3.9235118067106255e-06,
+      "loss": 0.1289,
+      "step": 291
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7682977344403001,
+      "learning_rate": 3.89190659586623e-06,
+      "loss": 0.1079,
+      "step": 292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8428528329835375,
+      "learning_rate": 3.8603478784017845e-06,
+      "loss": 0.1449,
+      "step": 293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4375467459626297,
+      "learning_rate": 3.828836978457868e-06,
+      "loss": 0.1042,
+      "step": 294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5484765703117677,
+      "learning_rate": 3.7973752181687336e-06,
+      "loss": 0.1316,
+      "step": 295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4447109885856044,
+      "learning_rate": 3.7659639176068287e-06,
+      "loss": 0.1002,
+      "step": 296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.591068651394467,
+      "learning_rate": 3.734604394727419e-06,
+      "loss": 0.1198,
+      "step": 297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5011976834408337,
+      "learning_rate": 3.703297965313275e-06,
+      "loss": 0.0882,
+      "step": 298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44482097536408105,
+      "learning_rate": 3.6720459429194743e-06,
+      "loss": 0.1019,
+      "step": 299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9176989963107355,
+      "learning_rate": 3.6408496388182857e-06,
+      "loss": 0.1691,
+      "step": 300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6875161477903405,
+      "learning_rate": 3.6097103619441505e-06,
+      "loss": 0.1277,
+      "step": 301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44943767486166497,
+      "learning_rate": 3.578629418838757e-06,
+      "loss": 0.1393,
+      "step": 302
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5447702492267633,
+      "learning_rate": 3.5476081135962335e-06,
+      "loss": 0.119,
+      "step": 303
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6759900156157556,
+      "learning_rate": 3.516647747808417e-06,
+      "loss": 0.1362,
+      "step": 304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4825645771030931,
+      "learning_rate": 3.4857496205102475e-06,
+      "loss": 0.1266,
+      "step": 305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5422813674696508,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.1163,
+      "step": 306
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5062407593658172,
+      "learning_rate": 3.4241452644112085e-06,
+      "loss": 0.1128,
+      "step": 307
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.555315224383365,
+      "learning_rate": 3.3934416204057396e-06,
+      "loss": 0.1282,
+      "step": 308
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6097420128434671,
+      "learning_rate": 3.3628053843722674e-06,
+      "loss": 0.1496,
+      "step": 309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5400763462235032,
+      "learning_rate": 3.3322378417458985e-06,
+      "loss": 0.1175,
+      "step": 310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5071327713034236,
+      "learning_rate": 3.3017402750794976e-06,
+      "loss": 0.0942,
+      "step": 311
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7040411771901197,
+      "learning_rate": 3.271313963989886e-06,
+      "loss": 0.1228,
+      "step": 312
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5394499701515973,
+      "learning_rate": 3.240960185104137e-06,
+      "loss": 0.1283,
+      "step": 313
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4069753524588952,
+      "learning_rate": 3.2106802120060197e-06,
+      "loss": 0.0953,
+      "step": 314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.43238651655204363,
+      "learning_rate": 3.180475315182563e-06,
+      "loss": 0.105,
+      "step": 315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6125882579505277,
+      "learning_rate": 3.1503467619707407e-06,
+      "loss": 0.1185,
+      "step": 316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3113829588086021,
+      "learning_rate": 3.1202958165043053e-06,
+      "loss": 0.1004,
+      "step": 317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38868095199448227,
+      "learning_rate": 3.090323739660742e-06,
+      "loss": 0.1018,
+      "step": 318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37965432285638184,
+      "learning_rate": 3.060431789008368e-06,
+      "loss": 0.0881,
+      "step": 319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.341947917294632,
+      "learning_rate": 3.0306212187535653e-06,
+      "loss": 0.0659,
+      "step": 320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.504882951287714,
+      "learning_rate": 3.000893279688155e-06,
+      "loss": 0.1127,
+      "step": 321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0372711232468814,
+      "learning_rate": 2.9712492191369245e-06,
+      "loss": 0.1296,
+      "step": 322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4063780308625291,
+      "learning_rate": 2.9416902809052817e-06,
+      "loss": 0.0989,
+      "step": 323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6151994077206249,
+      "learning_rate": 2.912217705227075e-06,
+      "loss": 0.1311,
+      "step": 324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29560996812170726,
+      "learning_rate": 2.882832728712551e-06,
+      "loss": 0.0826,
+      "step": 325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5419617814807206,
+      "learning_rate": 2.8535365842964713e-06,
+      "loss": 0.1031,
+      "step": 326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.316750361009068,
+      "learning_rate": 2.8243305011863843e-06,
+      "loss": 0.0859,
+      "step": 327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40660047128924554,
+      "learning_rate": 2.7952157048110406e-06,
+      "loss": 0.0972,
+      "step": 328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5248390219888603,
+      "learning_rate": 2.7661934167689887e-06,
+      "loss": 0.0973,
+      "step": 329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5506120889804264,
+      "learning_rate": 2.7372648547773063e-06,
+      "loss": 0.1358,
+      "step": 330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2933868543958218,
+      "learning_rate": 2.7084312326205164e-06,
+      "loss": 0.0665,
+      "step": 331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4260265900817017,
+      "learning_rate": 2.6796937600996587e-06,
+      "loss": 0.099,
+      "step": 332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5916472972429531,
+      "learning_rate": 2.6510536429815224e-06,
+      "loss": 0.1056,
+      "step": 333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25821980859434696,
+      "learning_rate": 2.622512082948063e-06,
+      "loss": 0.0826,
+      "step": 334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7602239496400257,
+      "learning_rate": 2.594070277545975e-06,
+      "loss": 0.1029,
+      "step": 335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3482619390273165,
+      "learning_rate": 2.5657294201364526e-06,
+      "loss": 0.1061,
+      "step": 336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5292694632995797,
+      "learning_rate": 2.5374906998451094e-06,
+      "loss": 0.1103,
+      "step": 337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.401776484530573,
+      "learning_rate": 2.5093553015120937e-06,
+      "loss": 0.1111,
+      "step": 338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33983736912957585,
+      "learning_rate": 2.4813244056423692e-06,
+      "loss": 0.1071,
+      "step": 339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3968757148636326,
+      "learning_rate": 2.4533991883561868e-06,
+      "loss": 0.0849,
+      "step": 340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37654025683974474,
+      "learning_rate": 2.425580821339733e-06,
+      "loss": 0.0699,
+      "step": 341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5455989773257783,
+      "learning_rate": 2.3978704717959777e-06,
+      "loss": 0.1098,
+      "step": 342
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8441196898407628,
+      "learning_rate": 2.3702693023956853e-06,
+      "loss": 0.1029,
+      "step": 343
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29473291180255373,
+      "learning_rate": 2.342778471228648e-06,
+      "loss": 0.0727,
+      "step": 344
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.43235629673761533,
+      "learning_rate": 2.315399131755081e-06,
+      "loss": 0.0923,
+      "step": 345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.548592005518116,
+      "learning_rate": 2.2881324327572336e-06,
+      "loss": 0.1064,
+      "step": 346
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3939095895213877,
+      "learning_rate": 2.260979518291186e-06,
+      "loss": 0.1241,
+      "step": 347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5355357896800356,
+      "learning_rate": 2.233941527638848e-06,
+      "loss": 0.097,
+      "step": 348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.365611233084578,
+      "learning_rate": 2.207019595260154e-06,
+      "loss": 0.0714,
+      "step": 349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4437291777264929,
+      "learning_rate": 2.1802148507454675e-06,
+      "loss": 0.0822,
+      "step": 350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35830740856943316,
+      "learning_rate": 2.1535284187681866e-06,
+      "loss": 0.1037,
+      "step": 351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4336882635339312,
+      "learning_rate": 2.1269614190375477e-06,
+      "loss": 0.1174,
+      "step": 352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6147280719205123,
+      "learning_rate": 2.1005149662516517e-06,
+      "loss": 0.1189,
+      "step": 353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32907167628090606,
+      "learning_rate": 2.07419017005069e-06,
+      "loss": 0.0879,
+      "step": 354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8879531426795192,
+      "learning_rate": 2.0479881349703885e-06,
+      "loss": 0.1018,
+      "step": 355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5784069701162355,
+      "learning_rate": 2.021909960395661e-06,
+      "loss": 0.0901,
+      "step": 356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4249807188430985,
+      "learning_rate": 1.9959567405144825e-06,
+      "loss": 0.1063,
+      "step": 357
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3740373486479186,
+      "learning_rate": 1.9701295642719836e-06,
+      "loss": 0.1055,
+      "step": 358
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31615505418225187,
+      "learning_rate": 1.944429515324749e-06,
+      "loss": 0.0829,
+      "step": 359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5605789927457013,
+      "learning_rate": 1.9188576719953635e-06,
+      "loss": 0.1123,
+      "step": 360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5206679107603537,
+      "learning_rate": 1.8934151072271573e-06,
+      "loss": 0.1262,
+      "step": 361
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32252663000252463,
+      "learning_rate": 1.8681028885391905e-06,
+      "loss": 0.079,
+      "step": 362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3906364665455239,
+      "learning_rate": 1.8429220779814654e-06,
+      "loss": 0.0948,
+      "step": 363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5027709696697109,
+      "learning_rate": 1.81787373209036e-06,
+      "loss": 0.1021,
+      "step": 364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3881070547527581,
+      "learning_rate": 1.7929589018443016e-06,
+      "loss": 0.0989,
+      "step": 365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3263559332241793,
+      "learning_rate": 1.7681786326196665e-06,
+      "loss": 0.0906,
+      "step": 366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.517125971430259,
+      "learning_rate": 1.743533964146924e-06,
+      "loss": 0.1249,
+      "step": 367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5577417175120104,
+      "learning_rate": 1.7190259304670038e-06,
+      "loss": 0.1146,
+      "step": 368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46739077120404987,
+      "learning_rate": 1.6946555598879138e-06,
+      "loss": 0.0787,
+      "step": 369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6368184328794871,
+      "learning_rate": 1.6704238749415958e-06,
+      "loss": 0.139,
+      "step": 370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4863805471570208,
+      "learning_rate": 1.6463318923410183e-06,
+      "loss": 0.0989,
+      "step": 371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5526864350993652,
+      "learning_rate": 1.6223806229375182e-06,
+      "loss": 0.1006,
+      "step": 372
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5042906795268511,
+      "learning_rate": 1.5985710716783936e-06,
+      "loss": 0.1254,
+      "step": 373
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5353859229799947,
+      "learning_rate": 1.5749042375647261e-06,
+      "loss": 0.129,
+      "step": 374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4983834015528933,
+      "learning_rate": 1.5513811136094786e-06,
+      "loss": 0.1219,
+      "step": 375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5736151839186128,
+      "learning_rate": 1.5280026867958186e-06,
+      "loss": 0.1052,
+      "step": 376
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4145349105653863,
+      "learning_rate": 1.5047699380357134e-06,
+      "loss": 0.105,
+      "step": 377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.522159341623551,
+      "learning_rate": 1.4816838421287693e-06,
+      "loss": 0.118,
+      "step": 378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6741011384826143,
+      "learning_rate": 1.4587453677213348e-06,
+      "loss": 0.104,
+      "step": 379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29249772565206017,
+      "learning_rate": 1.4359554772658551e-06,
+      "loss": 0.0859,
+      "step": 380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40320752444193486,
+      "learning_rate": 1.4133151269804873e-06,
+      "loss": 0.081,
+      "step": 381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3654144538640725,
+      "learning_rate": 1.39082526680899e-06,
+      "loss": 0.0904,
+      "step": 382
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37328450594396817,
+      "learning_rate": 1.368486840380851e-06,
+      "loss": 0.0959,
+      "step": 383
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3624280021978568,
+      "learning_rate": 1.3463007849717035e-06,
+      "loss": 0.0912,
+      "step": 384
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3236625582206281,
+      "learning_rate": 1.3242680314639995e-06,
+      "loss": 0.0815,
+      "step": 385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.50101809950523,
+      "learning_rate": 1.3023895043079476e-06,
+      "loss": 0.1347,
+      "step": 386
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41434448545848973,
+      "learning_rate": 1.2806661214827286e-06,
+      "loss": 0.0958,
+      "step": 387
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3791029114401193,
+      "learning_rate": 1.2590987944579808e-06,
+      "loss": 0.0849,
+      "step": 388
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4282672022245668,
+      "learning_rate": 1.2376884281555485e-06,
+      "loss": 0.1232,
+      "step": 389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30876155687450046,
+      "learning_rate": 1.2164359209115235e-06,
+      "loss": 0.0811,
+      "step": 390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5364966222557211,
+      "learning_rate": 1.1953421644385444e-06,
+      "loss": 0.1204,
+      "step": 391
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.506862615722067,
+      "learning_rate": 1.1744080437883859e-06,
+      "loss": 0.1185,
+      "step": 392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6111776038474351,
+      "learning_rate": 1.1536344373148245e-06,
+      "loss": 0.1482,
+      "step": 393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48878771363180995,
+      "learning_rate": 1.133022216636781e-06,
+      "loss": 0.105,
+      "step": 394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5675616899138232,
+      "learning_rate": 1.1125722466017547e-06,
+      "loss": 0.1113,
+      "step": 395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39125507848543734,
+      "learning_rate": 1.092285385249528e-06,
+      "loss": 0.107,
+      "step": 396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5831199696940993,
+      "learning_rate": 1.0721624837761768e-06,
+      "loss": 0.1293,
+      "step": 397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4427813558577841,
+      "learning_rate": 1.0522043864983428e-06,
+      "loss": 0.0956,
+      "step": 398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4969064755719846,
+      "learning_rate": 1.0324119308178166e-06,
+      "loss": 0.1219,
+      "step": 399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4177562502737496,
+      "learning_rate": 1.012785947186397e-06,
+      "loss": 0.1166,
+      "step": 400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3176953906595869,
+      "learning_rate": 9.933272590710508e-07,
+      "loss": 0.1044,
+      "step": 401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4819481702011284,
+      "learning_rate": 9.740366829193587e-07,
+      "loss": 0.1051,
+      "step": 402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40623333375228615,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.1169,
+      "step": 403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.35662447694983246,
+      "learning_rate": 9.359630969951012e-07,
+      "loss": 0.0933,
+      "step": 404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5222883713638226,
+      "learning_rate": 9.171816847139447e-07,
+      "loss": 0.1112,
+      "step": 405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2509639357115803,
+      "learning_rate": 8.985715793122407e-07,
+      "loss": 0.0599,
+      "step": 406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3369713146112169,
+      "learning_rate": 8.801335616327378e-07,
+      "loss": 0.0823,
+      "step": 407
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4162725251423473,
+      "learning_rate": 8.618684052977305e-07,
+      "loss": 0.0843,
+      "step": 408
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.637854600697931,
+      "learning_rate": 8.437768766765975e-07,
+      "loss": 0.1226,
+      "step": 409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4974824638764937,
+      "learning_rate": 8.258597348536452e-07,
+      "loss": 0.12,
+      "step": 410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4347267962714552,
+      "learning_rate": 8.081177315962601e-07,
+      "loss": 0.0779,
+      "step": 411
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6868660701065765,
+      "learning_rate": 7.905516113233652e-07,
+      "loss": 0.1269,
+      "step": 412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4881107761558822,
+      "learning_rate": 7.731621110741871e-07,
+      "loss": 0.1115,
+      "step": 413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.657770227545186,
+      "learning_rate": 7.55949960477328e-07,
+      "loss": 0.1442,
+      "step": 414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5169494480097382,
+      "learning_rate": 7.389158817201541e-07,
+      "loss": 0.135,
+      "step": 415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3181907269339257,
+      "learning_rate": 7.220605895184946e-07,
+      "loss": 0.0939,
+      "step": 416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32087685372237856,
+      "learning_rate": 7.053847910866513e-07,
+      "loss": 0.0798,
+      "step": 417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6806284443573277,
+      "learning_rate": 6.888891861077301e-07,
+      "loss": 0.1217,
+      "step": 418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46920132014349863,
+      "learning_rate": 6.725744667042778e-07,
+      "loss": 0.1143,
+      "step": 419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6856407917775545,
+      "learning_rate": 6.564413174092443e-07,
+      "loss": 0.0958,
+      "step": 420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24069693097745018,
+      "learning_rate": 6.404904151372649e-07,
+      "loss": 0.0593,
+      "step": 421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5917433333017736,
+      "learning_rate": 6.24722429156251e-07,
+      "loss": 0.1096,
+      "step": 422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1939796988122542,
+      "learning_rate": 6.091380210593145e-07,
+      "loss": 0.1099,
+      "step": 423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.596655388965177,
+      "learning_rate": 5.937378447370068e-07,
+      "loss": 0.1432,
+      "step": 424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.57879182871043,
+      "learning_rate": 5.785225463498828e-07,
+      "loss": 0.1327,
+      "step": 425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5509040715614761,
+      "learning_rate": 5.634927643013899e-07,
+      "loss": 0.1241,
+      "step": 426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7213197759630112,
+      "learning_rate": 5.486491292110796e-07,
+      "loss": 0.1698,
+      "step": 427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49287307205694036,
+      "learning_rate": 5.339922638881545e-07,
+      "loss": 0.1386,
+      "step": 428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3928406434495986,
+      "learning_rate": 5.195227833053273e-07,
+      "loss": 0.1031,
+      "step": 429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.372866049328715,
+      "learning_rate": 5.05241294573024e-07,
+      "loss": 0.0954,
+      "step": 430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.45727956067448033,
+      "learning_rate": 4.911483969139086e-07,
+      "loss": 0.1108,
+      "step": 431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5032913008903352,
+      "learning_rate": 4.772446816377408e-07,
+      "loss": 0.131,
+      "step": 432
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6426696510198964,
+      "learning_rate": 4.6353073211656886e-07,
+      "loss": 0.1052,
+      "step": 433
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5498263109499336,
+      "learning_rate": 4.5000712376024826e-07,
+      "loss": 0.1032,
+      "step": 434
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.40549938508345124,
+      "learning_rate": 4.3667442399229985e-07,
+      "loss": 0.0926,
+      "step": 435
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7541212490636288,
+      "learning_rate": 4.2353319222610265e-07,
+      "loss": 0.1439,
+      "step": 436
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4635173662535909,
+      "learning_rate": 4.1058397984142405e-07,
+      "loss": 0.096,
+      "step": 437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5998166471175362,
+      "learning_rate": 3.9782733016128006e-07,
+      "loss": 0.1109,
+      "step": 438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3380769566498276,
+      "learning_rate": 3.852637784291424e-07,
+      "loss": 0.0953,
+      "step": 439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38092780934292936,
+      "learning_rate": 3.728938517864794e-07,
+      "loss": 0.0945,
+      "step": 440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6024933875787805,
+      "learning_rate": 3.60718069250639e-07,
+      "loss": 0.1327,
+      "step": 441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44759338633310003,
+      "learning_rate": 3.4873694169306915e-07,
+      "loss": 0.0929,
+      "step": 442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5116789418410155,
+      "learning_rate": 3.369509718178887e-07,
+      "loss": 0.1275,
+      "step": 443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4017029932757541,
+      "learning_rate": 3.2536065414078724e-07,
+      "loss": 0.0855,
+      "step": 444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5016640455788635,
+      "learning_rate": 3.1396647496828245e-07,
+      "loss": 0.1004,
+      "step": 445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6339501144884603,
+      "learning_rate": 3.0276891237731085e-07,
+      "loss": 0.1266,
+      "step": 446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6132819225641576,
+      "learning_rate": 2.917684361951728e-07,
+      "loss": 0.1091,
+      "step": 447
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3294003954391178,
+      "learning_rate": 2.809655079798179e-07,
+      "loss": 0.0828,
+      "step": 448
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37041469935136134,
+      "learning_rate": 2.7036058100047723e-07,
+      "loss": 0.1038,
+      "step": 449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42103224834837355,
+      "learning_rate": 2.599541002186479e-07,
+      "loss": 0.1018,
+      "step": 450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.46321653430346443,
+      "learning_rate": 2.497465022694207e-07,
+      "loss": 0.1268,
+      "step": 451
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4537632438141447,
+      "learning_rate": 2.397382154431621e-07,
+      "loss": 0.1088,
+      "step": 452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7475463617224823,
+      "learning_rate": 2.2992965966754378e-07,
+      "loss": 0.1137,
+      "step": 453
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5796299422128173,
+      "learning_rate": 2.2032124648992015e-07,
+      "loss": 0.1336,
+      "step": 454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4691159453915289,
+      "learning_rate": 2.109133790600648e-07,
+      "loss": 0.108,
+      "step": 455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5657951237893031,
+      "learning_rate": 2.0170645211325335e-07,
+      "loss": 0.1171,
+      "step": 456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.47751387449511806,
+      "learning_rate": 1.9270085195370048e-07,
+      "loss": 0.1031,
+      "step": 457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3433510325579123,
+      "learning_rate": 1.838969564383525e-07,
+      "loss": 0.0892,
+      "step": 458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.638737100201716,
+      "learning_rate": 1.7529513496103322e-07,
+      "loss": 0.1385,
+      "step": 459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4782548614274575,
+      "learning_rate": 1.6689574843694433e-07,
+      "loss": 0.1234,
+      "step": 460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5326026090381726,
+      "learning_rate": 1.5869914928752117e-07,
+      "loss": 0.1189,
+      "step": 461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4669411712497723,
+      "learning_rate": 1.5070568142564912e-07,
+      "loss": 0.0891,
+      "step": 462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3862189370579595,
+      "learning_rate": 1.4291568024122848e-07,
+      "loss": 0.0831,
+      "step": 463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5448245008508469,
+      "learning_rate": 1.3532947258710905e-07,
+      "loss": 0.0955,
+      "step": 464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49220745304750935,
+      "learning_rate": 1.2794737676536993e-07,
+      "loss": 0.1051,
+      "step": 465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3637340784393114,
+      "learning_rate": 1.2076970251396593e-07,
+      "loss": 0.095,
+      "step": 466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3973355109883454,
+      "learning_rate": 1.1379675099373489e-07,
+      "loss": 0.1043,
+      "step": 467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44692984702888716,
+      "learning_rate": 1.0702881477575589e-07,
+      "loss": 0.1066,
+      "step": 468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5763679470176152,
+      "learning_rate": 1.004661778290783e-07,
+      "loss": 0.1234,
+      "step": 469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.44436077187137873,
+      "learning_rate": 9.410911550880474e-08,
+      "loss": 0.0918,
+      "step": 470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3476269676228987,
+      "learning_rate": 8.795789454453862e-08,
+      "loss": 0.0975,
+      "step": 471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46833587547634364,
+      "learning_rate": 8.201277302919086e-08,
+      "loss": 0.1242,
+      "step": 472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39017969110993045,
+      "learning_rate": 7.627400040815414e-08,
+      "loss": 0.101,
+      "step": 473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4768933617350448,
+      "learning_rate": 7.074181746883402e-08,
+      "loss": 0.0966,
+      "step": 474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3153236264194997,
+      "learning_rate": 6.54164563305465e-08,
+      "loss": 0.0978,
+      "step": 475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5166758551426724,
+      "learning_rate": 6.029814043478022e-08,
+      "loss": 0.0887,
+      "step": 476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3606207065519079,
+      "learning_rate": 5.538708453581787e-08,
+      "loss": 0.1031,
+      "step": 477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4529222503971012,
+      "learning_rate": 5.068349469173006e-08,
+      "loss": 0.1079,
+      "step": 478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5278299129516647,
+      "learning_rate": 4.618756825572612e-08,
+      "loss": 0.1106,
+      "step": 479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40312254426110233,
+      "learning_rate": 4.189949386787462e-08,
+      "loss": 0.0876,
+      "step": 480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42671161026738896,
+      "learning_rate": 3.781945144718912e-08,
+      "loss": 0.1113,
+      "step": 481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3887996362909368,
+      "learning_rate": 3.394761218407705e-08,
+      "loss": 0.094,
+      "step": 482
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37232757332940153,
+      "learning_rate": 3.0284138533160924e-08,
+      "loss": 0.106,
+      "step": 483
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3555996153166186,
+      "learning_rate": 2.6829184206457194e-08,
+      "loss": 0.0741,
+      "step": 484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.39075036068882,
+      "learning_rate": 2.358289416693027e-08,
+      "loss": 0.0997,
+      "step": 485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5388961437731601,
+      "learning_rate": 2.0545404622407396e-08,
+      "loss": 0.1184,
+      "step": 486
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4462529648205526,
+      "learning_rate": 1.7716843019867646e-08,
+      "loss": 0.1227,
+      "step": 487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7419559501017631,
+      "learning_rate": 1.509732804009012e-08,
+      "loss": 0.0997,
+      "step": 488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3574549728348087,
+      "learning_rate": 1.268696959267679e-08,
+      "loss": 0.0839,
+      "step": 489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7319674554514716,
+      "learning_rate": 1.0485868811441757e-08,
+      "loss": 0.1157,
+      "step": 490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4330086302002417,
+      "learning_rate": 8.494118050164646e-09,
+      "loss": 0.108,
+      "step": 491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.908790377545516,
+      "learning_rate": 6.711800878718144e-09,
+      "loss": 0.0784,
+      "step": 492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4162251131978262,
+      "learning_rate": 5.138992079561367e-09,
+      "loss": 0.1076,
+      "step": 493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6753032834413796,
+      "learning_rate": 3.775757644601808e-09,
+      "loss": 0.1421,
+      "step": 494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.351247714926159,
+      "learning_rate": 2.6221547724253337e-09,
+      "loss": 0.0838,
+      "step": 495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5659297756799074,
+      "learning_rate": 1.6782318658992159e-09,
+      "loss": 0.0941,
+      "step": 496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.579489706712821,
+      "learning_rate": 9.440285301370865e-10,
+      "loss": 0.0996,
+      "step": 497
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.42743719179238815,
+      "learning_rate": 4.1957557084082447e-10,
+      "loss": 0.1041,
+      "step": 498
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.37275044207396435,
+      "learning_rate": 1.0489499300603279e-10,
+      "loss": 0.0851,
+      "step": 499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5971765362779611,
+      "learning_rate": 0.0,
+      "loss": 0.1428,
+      "step": 500
+    },
+    {
+      "epoch": 1.0,
+      "step": 500,
+      "total_flos": 9092482900992.0,
+      "train_loss": 0.12780791262537242,
+      "train_runtime": 3709.5195,
+      "train_samples_per_second": 0.539,
+      "train_steps_per_second": 0.135
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "total_flos": 9092482900992.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_all_mm_tune_512_llm/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_all_mm_tune_512_llm/adapter_config.json

@@ -0,0 +1,28 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "k_proj",
+    "q_proj",
+    "fc1",
+    "up_proj",
+    "o_proj",
+    "v_proj",
+    "fc2",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_all_mm_tune_512_llm/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a0bfcb125505207db4fb326971c83434d4ef7cadb2076c77f156a87b5e749d6
+size 1529358010

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_all_mm_tune_512_llm/config.json

@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_all_mm_tune_512_llm/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_1_olora256_all_mm_tune_512_llm/non_lora_trainables.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd98657814ebbbc0ea316972b27ebeb379ff65dcea996e404b1bd66668d5f9f5
+size 33964208

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_all_mm_tune_olora256_512_llm/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_all_mm_tune_olora256_512_llm/adapter_config.json

@@ -0,0 +1,28 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "v_proj",
+    "o_proj",
+    "fc1",
+    "k_proj",
+    "fc2",
+    "up_proj",
+    "q_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_all_mm_tune_olora256_512_llm/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d2f4f19e7626d066ae39da23c434add781c82ed522d560c25d4bcdc09c4d5cc
+size 1529358010

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_all_mm_tune_olora256_512_llm/config.json

@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_all_mm_tune_olora256_512_llm/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_all_mm_tune_olora256_512_llm/non_lora_trainables.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:501122b5a1df3a7cbf0688fdcfa42ddc89cbf1aaef2ef29e1bce5ba1a5b95f89
+size 33964208

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_bash_olora256_512_llm/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_bash_olora256_512_llm/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "k_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "q_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_bash_olora256_512_llm/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8836bd393ca746e68ac2d18d9c18d7eca982afe5880b1af6af5811693691cf40
+size 1384057050

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_bash_olora256_512_llm/config.json

@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_bash_olora256_512_llm/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_23_bash_olora256_512_llm/non_lora_trainables.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
+size 912

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_512_llm/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_512_llm/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "v_proj",
+    "down_proj",
+    "o_proj",
+    "k_proj",
+    "gate_proj",
+    "q_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_512_llm/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40a49694ef48a189c1ff9aa59a62850e1552f0b0531610ef54943c480d20910a
+size 1384057050

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_512_llm/config.json

@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_512_llm/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_512_llm/non_lora_trainables.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
+size 912

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_all_mm_tune_512_llm/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_all_mm_tune_512_llm/adapter_config.json

@@ -0,0 +1,28 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "fc2",
+    "q_proj",
+    "k_proj",
+    "gate_proj",
+    "v_proj",
+    "fc1",
+    "up_proj",
+    "o_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_all_mm_tune_512_llm/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5ceb3e92a6af5268286c0d3052f07c60edc638610da316c97a3e7917dcee8ae
+size 1529358010

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_all_mm_tune_512_llm/config.json

@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_all_mm_tune_512_llm/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_2_olora256_all_mm_tune_512_llm/non_lora_trainables.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f85c885a28521bd93b4926f40e5fbf657d431bab27a2caea5c4baa5154e3e672
+size 33964208

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_6_olora256_512_llm/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_6_olora256_512_llm/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": "olora",
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 512,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "v_proj",
+    "gate_proj",
+    "q_proj",
+    "k_proj",
+    "up_proj",
+    "o_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_6_olora256_512_llm/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d44cde697b4a72164e7fe3957f580825c8bb8cfc341a347a98574a7aee90c641
+size 1384057050

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_6_olora256_512_llm/config.json

@@ -0,0 +1,221 @@
+{
+  "_name_or_path": "lmms-lab/llava-onevision-qwen2-7b-si",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "faster_token_stride": 10,
+  "force_sample": true,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_spatial_pool_stride": 2,
+  "mm_tunable_parts": "mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": null,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.0.dev0",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_6_olora256_512_llm/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.40.0.dev0"
+}

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_6_olora256_512_llm/non_lora_trainables.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
+size 912

llavanext-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split0_6_olora256_all_mm_tune_512_llm/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.4.0