Training in progress, epoch 12, checkpoint

checkpoint-3000/README.md

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

checkpoint-3000/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-3000/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95d652a4824b2323af280f678840ceecbec5674ab7c8f4c684c2592468be675e
+size 112130416

checkpoint-3000/added_tokens.json

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

checkpoint-3000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8077aaaf35d9d1849827d934e5eb835b3964d85d1d7fa7d75836bce54355d3d2
+size 224537202

checkpoint-3000/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20e3fee63f623c80252a3bdd070c843ec7b1a0e5e7c405652a434ee209a6fd87
+size 15024

checkpoint-3000/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbd8e7bb81fb3287d85f2d25e23d2759b46d3c8152c5cfffa5f93ff4c62e7e8b
+size 15024

checkpoint-3000/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:275a098d43c6310bcab52eabe282fb209065e6b67673726dc2077ed4afdcf560
+size 15024

checkpoint-3000/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e65685127b76e246a04543812c2bfcef68fb8adb65d1eb3fe43b15bb182ea173
+size 15024

checkpoint-3000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61725e453643aca3b98d3f6fe59cc89762ac80a6e72955656f8de7c9fa16e931
+size 1064

checkpoint-3000/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-3000/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-3000/trainer_state.json

@@ -0,0 +1,423 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 12.0,
+  "eval_steps": 500,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5897082090377808,
+      "learning_rate": 0.0001,
+      "loss": 2.3566,
+      "step": 100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5390245914459229,
+      "learning_rate": 0.0001,
+      "loss": 2.2503,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5155873015873016,
+      "eval_loss": 2.3237216472625732,
+      "eval_runtime": 6.3164,
+      "eval_samples_per_second": 79.159,
+      "eval_steps_per_second": 9.974,
+      "step": 250
+    },
+    {
+      "epoch": 1.0,
+      "eval_exact_match": 17.2,
+      "eval_f1": 25.586581359816655,
+      "step": 250
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7468857169151306,
+      "learning_rate": 0.0001,
+      "loss": 2.1798,
+      "step": 300
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9109599590301514,
+      "learning_rate": 0.0001,
+      "loss": 2.1094,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8458774089813232,
+      "learning_rate": 0.0001,
+      "loss": 2.087,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5163809523809524,
+      "eval_loss": 2.3309242725372314,
+      "eval_runtime": 5.1896,
+      "eval_samples_per_second": 96.347,
+      "eval_steps_per_second": 12.14,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 18.6,
+      "eval_f1": 26.043196248196264,
+      "step": 500
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.171212077140808,
+      "learning_rate": 0.0001,
+      "loss": 1.8413,
+      "step": 600
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.1294909715652466,
+      "learning_rate": 0.0001,
+      "loss": 1.849,
+      "step": 700
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5144761904761905,
+      "eval_loss": 2.401942253112793,
+      "eval_runtime": 5.5775,
+      "eval_samples_per_second": 89.646,
+      "eval_steps_per_second": 11.295,
+      "step": 750
+    },
+    {
+      "epoch": 3.0,
+      "eval_exact_match": 18.4,
+      "eval_f1": 26.966370002546494,
+      "step": 750
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.5013824701309204,
+      "learning_rate": 0.0001,
+      "loss": 1.7322,
+      "step": 800
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.4516321420669556,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 900
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.6037466526031494,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5103809523809524,
+      "eval_loss": 2.5039021968841553,
+      "eval_runtime": 5.1568,
+      "eval_samples_per_second": 96.96,
+      "eval_steps_per_second": 12.217,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 18.4,
+      "eval_f1": 26.892813852813866,
+      "step": 1000
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.7909170389175415,
+      "learning_rate": 0.0001,
+      "loss": 1.3298,
+      "step": 1100
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.9236161708831787,
+      "learning_rate": 0.0001,
+      "loss": 1.3666,
+      "step": 1200
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.505015873015873,
+      "eval_loss": 2.6543924808502197,
+      "eval_runtime": 5.5115,
+      "eval_samples_per_second": 90.719,
+      "eval_steps_per_second": 11.431,
+      "step": 1250
+    },
+    {
+      "epoch": 5.0,
+      "eval_exact_match": 18.4,
+      "eval_f1": 25.875286609941792,
+      "step": 1250
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 1.8524874448776245,
+      "learning_rate": 0.0001,
+      "loss": 1.2381,
+      "step": 1300
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 2.1092071533203125,
+      "learning_rate": 0.0001,
+      "loss": 1.1141,
+      "step": 1400
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.065767288208008,
+      "learning_rate": 0.0001,
+      "loss": 1.1435,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5010793650793651,
+      "eval_loss": 2.843566656112671,
+      "eval_runtime": 5.1394,
+      "eval_samples_per_second": 97.287,
+      "eval_steps_per_second": 12.258,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 17.4,
+      "eval_f1": 24.31514874014875,
+      "step": 1500
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 2.329406261444092,
+      "learning_rate": 0.0001,
+      "loss": 0.8887,
+      "step": 1600
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.489598274230957,
+      "learning_rate": 0.0001,
+      "loss": 0.9171,
+      "step": 1700
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.49707936507936507,
+      "eval_loss": 3.0319864749908447,
+      "eval_runtime": 5.3015,
+      "eval_samples_per_second": 94.313,
+      "eval_steps_per_second": 11.883,
+      "step": 1750
+    },
+    {
+      "epoch": 7.0,
+      "eval_exact_match": 16.0,
+      "eval_f1": 22.824985569985575,
+      "step": 1750
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 2.764033794403076,
+      "learning_rate": 0.0001,
+      "loss": 0.8093,
+      "step": 1800
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 2.5643441677093506,
+      "learning_rate": 0.0001,
+      "loss": 0.724,
+      "step": 1900
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.074833631515503,
+      "learning_rate": 0.0001,
+      "loss": 0.7531,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.49304761904761907,
+      "eval_loss": 3.2584712505340576,
+      "eval_runtime": 5.2945,
+      "eval_samples_per_second": 94.438,
+      "eval_steps_per_second": 11.899,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 15.0,
+      "eval_f1": 21.895264180264185,
+      "step": 2000
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 2.649679660797119,
+      "learning_rate": 0.0001,
+      "loss": 0.5777,
+      "step": 2100
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 2.1969234943389893,
+      "learning_rate": 0.0001,
+      "loss": 0.6101,
+      "step": 2200
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.4925079365079365,
+      "eval_loss": 3.3418095111846924,
+      "eval_runtime": 5.1381,
+      "eval_samples_per_second": 97.312,
+      "eval_steps_per_second": 12.261,
+      "step": 2250
+    },
+    {
+      "epoch": 9.0,
+      "eval_exact_match": 14.4,
+      "eval_f1": 21.76202020202021,
+      "step": 2250
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 1.8318747282028198,
+      "learning_rate": 0.0001,
+      "loss": 0.555,
+      "step": 2300
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 1.9915804862976074,
+      "learning_rate": 0.0001,
+      "loss": 0.5102,
+      "step": 2400
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 2.1287789344787598,
+      "learning_rate": 0.0001,
+      "loss": 0.5392,
+      "step": 2500
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.4915555555555556,
+      "eval_loss": 3.537275791168213,
+      "eval_runtime": 5.1264,
+      "eval_samples_per_second": 97.534,
+      "eval_steps_per_second": 12.289,
+      "step": 2500
+    },
+    {
+      "epoch": 10.0,
+      "eval_exact_match": 15.4,
+      "eval_f1": 22.23013114336644,
+      "step": 2500
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 1.6748021841049194,
+      "learning_rate": 0.0001,
+      "loss": 0.4426,
+      "step": 2600
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 1.8744910955429077,
+      "learning_rate": 0.0001,
+      "loss": 0.4718,
+      "step": 2700
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.48926984126984124,
+      "eval_loss": 3.631282091140747,
+      "eval_runtime": 5.3142,
+      "eval_samples_per_second": 94.087,
+      "eval_steps_per_second": 11.855,
+      "step": 2750
+    },
+    {
+      "epoch": 11.0,
+      "eval_exact_match": 15.6,
+      "eval_f1": 21.87386002886003,
+      "step": 2750
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 2.1398115158081055,
+      "learning_rate": 0.0001,
+      "loss": 0.4456,
+      "step": 2800
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 1.8851490020751953,
+      "learning_rate": 0.0001,
+      "loss": 0.4328,
+      "step": 2900
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 1.6958885192871094,
+      "learning_rate": 0.0001,
+      "loss": 0.4446,
+      "step": 3000
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.49057142857142855,
+      "eval_loss": 3.6736366748809814,
+      "eval_runtime": 5.8652,
+      "eval_samples_per_second": 85.248,
+      "eval_steps_per_second": 10.741,
+      "step": 3000
+    },
+    {
+      "epoch": 12.0,
+      "eval_exact_match": 15.2,
+      "eval_f1": 21.81304966962862,
+      "step": 3000
+    }
+  ],
+  "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": 9.178062811404698e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-3000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3414153f69372b550d629c900d05ab7510b14d218f2d8546142e7b41d948d3e
+size 5304