Training in progress, epoch 9, checkpoint

checkpoint-2250/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.5.0

checkpoint-2250/adapter_config.json

@@ -0,0 +1,25 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen1.5-4B",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "k_proj",
+    "v_proj",
+    "down_proj",
+    "gate_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-2250/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a95573aba8c7ad85c359a1780e46d35f907c374ea6020f62c523b0d4458357dd
+size 112130416

checkpoint-2250/added_tokens.json

@@ -0,0 +1,5 @@
+{
+  "<|endoftext|>": 151643,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644
+}

checkpoint-2250/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f54038663adf649aef58e255d91d67b80bad398c9879392bbf9eac2aba0f5d7
+size 224537202

checkpoint-2250/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10ad9d3c8da575eab4dce1d0fd27faae8881ed2f8b7d824ea91554972be14202
+size 15024

checkpoint-2250/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc2e1e7c7596fa8ac9a427794b880699469ff72097fa6a7045634b6b2a1944eb
+size 15024

checkpoint-2250/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b5c3866d0d8c1601c3bbe0f09e3030991510b4b05cd9d8334e676828f55455e
+size 15024

checkpoint-2250/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcbaf3716a263b8a7d4f7d20c10c643931b00647e2ea4628fb6fa5b762283d73
+size 15024

checkpoint-2250/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7247d6ea2dea3e7cc89e3cddfa6c19967e39774276554ee2f3b8ddaa2530ced
+size 1064

checkpoint-2250/special_tokens_map.json

@@ -0,0 +1,14 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|endoftext|>"
+}

checkpoint-2250/tokenizer_config.json

@@ -0,0 +1,43 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-2250/trainer_state.json

@@ -0,0 +1,322 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.0,
+  "eval_steps": 500,
+  "global_step": 2250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6831097602844238,
+      "learning_rate": 3e-05,
+      "loss": 2.487,
+      "step": 100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7073227763175964,
+      "learning_rate": 3e-05,
+      "loss": 2.276,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5134285714285715,
+      "eval_loss": 2.3502681255340576,
+      "eval_runtime": 5.2185,
+      "eval_samples_per_second": 95.812,
+      "eval_steps_per_second": 12.072,
+      "step": 250
+    },
+    {
+      "epoch": 1.0,
+      "eval_exact_match": 17.0,
+      "eval_f1": 24.54219480519481,
+      "step": 250
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.820932149887085,
+      "learning_rate": 3e-05,
+      "loss": 2.2479,
+      "step": 300
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9243379831314087,
+      "learning_rate": 3e-05,
+      "loss": 2.2237,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8742844462394714,
+      "learning_rate": 3e-05,
+      "loss": 2.1973,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5151746031746032,
+      "eval_loss": 2.330955743789673,
+      "eval_runtime": 5.5478,
+      "eval_samples_per_second": 90.125,
+      "eval_steps_per_second": 11.356,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 16.2,
+      "eval_f1": 24.397655122655127,
+      "step": 500
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.034700632095337,
+      "learning_rate": 3e-05,
+      "loss": 2.1295,
+      "step": 600
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.130967617034912,
+      "learning_rate": 3e-05,
+      "loss": 2.1128,
+      "step": 700
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.516095238095238,
+      "eval_loss": 2.3357651233673096,
+      "eval_runtime": 5.8761,
+      "eval_samples_per_second": 85.091,
+      "eval_steps_per_second": 10.721,
+      "step": 750
+    },
+    {
+      "epoch": 3.0,
+      "eval_exact_match": 19.0,
+      "eval_f1": 26.202922077922086,
+      "step": 750
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.4410409927368164,
+      "learning_rate": 3e-05,
+      "loss": 2.0768,
+      "step": 800
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.5838419198989868,
+      "learning_rate": 3e-05,
+      "loss": 2.0139,
+      "step": 900
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.824843406677246,
+      "learning_rate": 3e-05,
+      "loss": 2.0165,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5151746031746032,
+      "eval_loss": 2.355579376220703,
+      "eval_runtime": 5.1124,
+      "eval_samples_per_second": 97.801,
+      "eval_steps_per_second": 12.323,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 19.0,
+      "eval_f1": 26.60538961038962,
+      "step": 1000
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.988659381866455,
+      "learning_rate": 3e-05,
+      "loss": 1.8919,
+      "step": 1100
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 2.1034774780273438,
+      "learning_rate": 3e-05,
+      "loss": 1.9024,
+      "step": 1200
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5138730158730159,
+      "eval_loss": 2.4047114849090576,
+      "eval_runtime": 5.507,
+      "eval_samples_per_second": 90.794,
+      "eval_steps_per_second": 11.44,
+      "step": 1250
+    },
+    {
+      "epoch": 5.0,
+      "eval_exact_match": 19.2,
+      "eval_f1": 27.08811688311689,
+      "step": 1250
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 2.2108240127563477,
+      "learning_rate": 3e-05,
+      "loss": 1.852,
+      "step": 1300
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 2.507085084915161,
+      "learning_rate": 3e-05,
+      "loss": 1.7877,
+      "step": 1400
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.3765366077423096,
+      "learning_rate": 3e-05,
+      "loss": 1.7989,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5111111111111111,
+      "eval_loss": 2.4534740447998047,
+      "eval_runtime": 5.199,
+      "eval_samples_per_second": 96.172,
+      "eval_steps_per_second": 12.118,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 18.2,
+      "eval_f1": 26.325503531556176,
+      "step": 1500
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 2.944777727127075,
+      "learning_rate": 3e-05,
+      "loss": 1.6784,
+      "step": 1600
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.8724122047424316,
+      "learning_rate": 3e-05,
+      "loss": 1.692,
+      "step": 1700
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5085714285714286,
+      "eval_loss": 2.527308464050293,
+      "eval_runtime": 5.9179,
+      "eval_samples_per_second": 84.489,
+      "eval_steps_per_second": 10.646,
+      "step": 1750
+    },
+    {
+      "epoch": 7.0,
+      "eval_exact_match": 17.4,
+      "eval_f1": 25.36568542568543,
+      "step": 1750
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 3.7004809379577637,
+      "learning_rate": 3e-05,
+      "loss": 1.6336,
+      "step": 1800
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 3.611534595489502,
+      "learning_rate": 3e-05,
+      "loss": 1.5772,
+      "step": 1900
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.311814546585083,
+      "learning_rate": 3e-05,
+      "loss": 1.5963,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5057142857142857,
+      "eval_loss": 2.6106724739074707,
+      "eval_runtime": 5.9096,
+      "eval_samples_per_second": 84.608,
+      "eval_steps_per_second": 10.661,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 16.4,
+      "eval_f1": 24.649942279942284,
+      "step": 2000
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 4.057112693786621,
+      "learning_rate": 3e-05,
+      "loss": 1.4604,
+      "step": 2100
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 3.5819571018218994,
+      "learning_rate": 3e-05,
+      "loss": 1.482,
+      "step": 2200
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.5044761904761905,
+      "eval_loss": 2.6719970703125,
+      "eval_runtime": 5.8123,
+      "eval_samples_per_second": 86.024,
+      "eval_steps_per_second": 10.839,
+      "step": 2250
+    },
+    {
+      "epoch": 9.0,
+      "eval_exact_match": 16.0,
+      "eval_f1": 24.24823953823954,
+      "step": 2250
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 12500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.881574399875482e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-2250/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76141cb7041782161f7695865881a4df3a03be3731ae75640f405be3a5a6f23d
+size 5304