Training in progress, epoch 12, 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 sha256:775a8680a997346df04e29b2c23988979da49ed4731f9bd8a2972fa59d7ec41c
+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:eee5eb5a2f087e712320584fb569870f32891584d6e926eb6cf0e3d785e02bfb
+size 224537202

checkpoint-2250/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bed9e0e768857ac5730d4e7ca22bd300d942024e893658a4faaa1bfd775f045
+size 15024

checkpoint-2250/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0692bd6406e4f63273a4d6df520cf10465618ab93b7581cd4a0ec4c411f9dd17
+size 15024

checkpoint-2250/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0965155fdb7f3b400d5063281d8712d30459327808a911ceec296723d6cd55e4
+size 15024

checkpoint-2250/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e419d52ed190f081567d82312a2578887d7cc86ab10a458ff4ddbb69b0e18ca
+size 15024

checkpoint-2250/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9632c9f0d294c00937b894a7699a18d609a94659621a0675d7649288ad07f95
+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 +1,367 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 12.0,
+  "eval_steps": 500,
+  "global_step": 2250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.7454617619514465,
+      "learning_rate": 5e-05,
+      "loss": 1.8726,
+      "step": 100
+    },
+    {
+      "epoch": 0.9973333333333333,
+      "eval_accuracy": 0.5713333333333334,
+      "eval_loss": 1.6331332921981812,
+      "eval_runtime": 5.4718,
+      "eval_samples_per_second": 91.377,
+      "eval_steps_per_second": 11.514,
+      "step": 187
+    },
+    {
+      "epoch": 0.9973333333333333,
+      "eval_exact_match": 15.0,
+      "eval_f1": 25.64857142857145,
+      "step": 187
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 0.9573808908462524,
+      "learning_rate": 5e-05,
+      "loss": 1.66,
+      "step": 200
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.0792433023452759,
+      "learning_rate": 5e-05,
+      "loss": 1.5764,
+      "step": 300
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5741538461538461,
+      "eval_loss": 1.6118417978286743,
+      "eval_runtime": 5.1541,
+      "eval_samples_per_second": 97.01,
+      "eval_steps_per_second": 12.223,
+      "step": 375
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 16.8,
+      "eval_f1": 26.120952380952396,
+      "step": 375
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 1.2928162813186646,
+      "learning_rate": 5e-05,
+      "loss": 1.5364,
+      "step": 400
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 1.4883633852005005,
+      "learning_rate": 5e-05,
+      "loss": 1.4154,
+      "step": 500
+    },
+    {
+      "epoch": 2.997333333333333,
+      "eval_accuracy": 0.5733846153846154,
+      "eval_loss": 1.6421321630477905,
+      "eval_runtime": 5.23,
+      "eval_samples_per_second": 95.603,
+      "eval_steps_per_second": 12.046,
+      "step": 562
+    },
+    {
+      "epoch": 2.997333333333333,
+      "eval_exact_match": 19.4,
+      "eval_f1": 28.4652380952381,
+      "step": 562
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.7553108930587769,
+      "learning_rate": 5e-05,
+      "loss": 1.3426,
+      "step": 600
+    },
+    {
+      "epoch": 3.7333333333333334,
+      "grad_norm": 2.0343973636627197,
+      "learning_rate": 5e-05,
+      "loss": 1.2297,
+      "step": 700
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5687692307692308,
+      "eval_loss": 1.7009365558624268,
+      "eval_runtime": 5.4367,
+      "eval_samples_per_second": 91.968,
+      "eval_steps_per_second": 11.588,
+      "step": 750
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 17.6,
+      "eval_f1": 26.490952380952383,
+      "step": 750
+    },
+    {
+      "epoch": 4.266666666666667,
+      "grad_norm": 2.3617172241210938,
+      "learning_rate": 5e-05,
+      "loss": 1.1613,
+      "step": 800
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 2.357593297958374,
+      "learning_rate": 5e-05,
+      "loss": 1.0779,
+      "step": 900
+    },
+    {
+      "epoch": 4.997333333333334,
+      "eval_accuracy": 0.567025641025641,
+      "eval_loss": 1.791063666343689,
+      "eval_runtime": 6.0902,
+      "eval_samples_per_second": 82.099,
+      "eval_steps_per_second": 10.345,
+      "step": 937
+    },
+    {
+      "epoch": 4.997333333333334,
+      "eval_exact_match": 16.0,
+      "eval_f1": 23.740000000000013,
+      "step": 937
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 2.7846109867095947,
+      "learning_rate": 5e-05,
+      "loss": 0.9807,
+      "step": 1000
+    },
+    {
+      "epoch": 5.866666666666667,
+      "grad_norm": 2.713942050933838,
+      "learning_rate": 5e-05,
+      "loss": 0.9547,
+      "step": 1100
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.563076923076923,
+      "eval_loss": 1.8913958072662354,
+      "eval_runtime": 5.2006,
+      "eval_samples_per_second": 96.142,
+      "eval_steps_per_second": 12.114,
+      "step": 1125
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 15.6,
+      "eval_f1": 24.21666666666668,
+      "step": 1125
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 2.6644272804260254,
+      "learning_rate": 5e-05,
+      "loss": 0.8401,
+      "step": 1200
+    },
+    {
+      "epoch": 6.933333333333334,
+      "grad_norm": 3.0467491149902344,
+      "learning_rate": 5e-05,
+      "loss": 0.8414,
+      "step": 1300
+    },
+    {
+      "epoch": 6.997333333333334,
+      "eval_accuracy": 0.5606666666666666,
+      "eval_loss": 1.987471342086792,
+      "eval_runtime": 5.1815,
+      "eval_samples_per_second": 96.497,
+      "eval_steps_per_second": 12.159,
+      "step": 1312
+    },
+    {
+      "epoch": 6.997333333333334,
+      "eval_exact_match": 15.4,
+      "eval_f1": 24.453174603174613,
+      "step": 1312
+    },
+    {
+      "epoch": 7.466666666666667,
+      "grad_norm": 2.872312545776367,
+      "learning_rate": 5e-05,
+      "loss": 0.7353,
+      "step": 1400
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.138514995574951,
+      "learning_rate": 5e-05,
+      "loss": 0.7568,
+      "step": 1500
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5592820512820513,
+      "eval_loss": 2.061145782470703,
+      "eval_runtime": 5.0637,
+      "eval_samples_per_second": 98.741,
+      "eval_steps_per_second": 12.441,
+      "step": 1500
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 15.4,
+      "eval_f1": 23.31000000000001,
+      "step": 1500
+    },
+    {
+      "epoch": 8.533333333333333,
+      "grad_norm": 2.4859418869018555,
+      "learning_rate": 5e-05,
+      "loss": 0.6477,
+      "step": 1600
+    },
+    {
+      "epoch": 8.997333333333334,
+      "eval_accuracy": 0.555025641025641,
+      "eval_loss": 2.1961007118225098,
+      "eval_runtime": 5.166,
+      "eval_samples_per_second": 96.788,
+      "eval_steps_per_second": 12.195,
+      "step": 1687
+    },
+    {
+      "epoch": 8.997333333333334,
+      "eval_exact_match": 14.8,
+      "eval_f1": 22.49904761904763,
+      "step": 1687
+    },
+    {
+      "epoch": 9.066666666666666,
+      "grad_norm": 2.5425331592559814,
+      "learning_rate": 5e-05,
+      "loss": 0.6783,
+      "step": 1700
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 2.923581838607788,
+      "learning_rate": 5e-05,
+      "loss": 0.6137,
+      "step": 1800
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.554923076923077,
+      "eval_loss": 2.2456719875335693,
+      "eval_runtime": 5.2657,
+      "eval_samples_per_second": 94.955,
+      "eval_steps_per_second": 11.964,
+      "step": 1875
+    },
+    {
+      "epoch": 10.0,
+      "eval_exact_match": 14.4,
+      "eval_f1": 22.522857142857152,
+      "step": 1875
+    },
+    {
+      "epoch": 10.133333333333333,
+      "grad_norm": 2.56027889251709,
+      "learning_rate": 5e-05,
+      "loss": 0.6148,
+      "step": 1900
+    },
+    {
+      "epoch": 10.666666666666666,
+      "grad_norm": 2.515385389328003,
+      "learning_rate": 5e-05,
+      "loss": 0.5823,
+      "step": 2000
+    },
+    {
+      "epoch": 10.997333333333334,
+      "eval_accuracy": 0.5546666666666666,
+      "eval_loss": 2.296849012374878,
+      "eval_runtime": 5.8609,
+      "eval_samples_per_second": 85.311,
+      "eval_steps_per_second": 10.749,
+      "step": 2062
+    },
+    {
+      "epoch": 10.997333333333334,
+      "eval_exact_match": 14.6,
+      "eval_f1": 23.82142857142859,
+      "step": 2062
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 2.405318021774292,
+      "learning_rate": 5e-05,
+      "loss": 0.5838,
+      "step": 2100
+    },
+    {
+      "epoch": 11.733333333333333,
+      "grad_norm": 2.2395944595336914,
+      "learning_rate": 5e-05,
+      "loss": 0.5661,
+      "step": 2200
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.5551794871794872,
+      "eval_loss": 2.3467304706573486,
+      "eval_runtime": 5.1088,
+      "eval_samples_per_second": 97.87,
+      "eval_steps_per_second": 12.332,
+      "step": 2250
+    },
+    {
+      "epoch": 12.0,
+      "eval_exact_match": 14.8,
+      "eval_f1": 23.336190476190485,
+      "step": 2250
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 9350,
+  "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": 3.918554630769869e+16,
+  "train_batch_size": 1,
+  "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:dd2191b7045c2a2985850d3b9f1472eaf3359ca38b9bc68efae9970e68b817ab
+size 5304