Training in progress, epoch 10, checkpoint

checkpoint-2500/README.md

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

checkpoint-2500/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-2500/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fe07b494ea072de1220c438488b756b7e06d0f88c2853b9680a038f4a3c579d
+size 112130416

checkpoint-2500/added_tokens.json

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

checkpoint-2500/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b034b6eda68e14754cb7a0b53da5756a79bc62cda5b637c4419c431a1dd72db5
+size 224537202

checkpoint-2500/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15ce3df7625f5f544fc5980ebde5bb0e2c36f876568d674999952ac9235dc223
+size 15024

checkpoint-2500/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62e630ac9aa2363da8f452192ac54ad4741aa66cebe9feba639f8ea64bc4c9ec
+size 15024

checkpoint-2500/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e3aaec5208d1c4caaff694030ecbf24896f9c59110edde3a65afbd8dafbdabb
+size 15024

checkpoint-2500/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94c26bd3362efc405de762229b02da7cfc5538bf190f87c0b649472186c2b5a5
+size 15024

checkpoint-2500/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9de07d7293968b9d2922a7f8c0991f17ddef6d3caf4a35f0fc3606f616dc368
+size 1064

checkpoint-2500/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-2500/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-2500/trainer_state.json

@@ -0,0 +1,358 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6251392960548401,
+      "learning_rate": 5e-05,
+      "loss": 2.427,
+      "step": 100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6590011715888977,
+      "learning_rate": 5e-05,
+      "loss": 2.2748,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5142539682539683,
+      "eval_loss": 2.336690902709961,
+      "eval_runtime": 5.4733,
+      "eval_samples_per_second": 91.353,
+      "eval_steps_per_second": 11.51,
+      "step": 250
+    },
+    {
+      "epoch": 1.0,
+      "eval_exact_match": 16.8,
+      "eval_f1": 24.850246583481884,
+      "step": 250
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8194239139556885,
+      "learning_rate": 5e-05,
+      "loss": 2.2322,
+      "step": 300
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9867294430732727,
+      "learning_rate": 5e-05,
+      "loss": 2.1904,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9660400152206421,
+      "learning_rate": 5e-05,
+      "loss": 2.1657,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5168253968253969,
+      "eval_loss": 2.323688507080078,
+      "eval_runtime": 5.1683,
+      "eval_samples_per_second": 96.744,
+      "eval_steps_per_second": 12.19,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 17.2,
+      "eval_f1": 24.620036075036083,
+      "step": 500
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.281128168106079,
+      "learning_rate": 5e-05,
+      "loss": 2.0251,
+      "step": 600
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.3365116119384766,
+      "learning_rate": 5e-05,
+      "loss": 2.0158,
+      "step": 700
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5156507936507937,
+      "eval_loss": 2.3560824394226074,
+      "eval_runtime": 5.4675,
+      "eval_samples_per_second": 91.449,
+      "eval_steps_per_second": 11.523,
+      "step": 750
+    },
+    {
+      "epoch": 3.0,
+      "eval_exact_match": 19.0,
+      "eval_f1": 27.242790688843332,
+      "step": 750
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.7840828895568848,
+      "learning_rate": 5e-05,
+      "loss": 1.9394,
+      "step": 800
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.8030190467834473,
+      "learning_rate": 5e-05,
+      "loss": 1.8502,
+      "step": 900
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.9682222604751587,
+      "learning_rate": 5e-05,
+      "loss": 1.8682,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5131746031746032,
+      "eval_loss": 2.4071946144104004,
+      "eval_runtime": 5.2825,
+      "eval_samples_per_second": 94.653,
+      "eval_steps_per_second": 11.926,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 17.6,
+      "eval_f1": 26.010988455988468,
+      "step": 1000
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 2.138505220413208,
+      "learning_rate": 5e-05,
+      "loss": 1.6795,
+      "step": 1100
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 2.3212926387786865,
+      "learning_rate": 5e-05,
+      "loss": 1.7036,
+      "step": 1200
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5099365079365079,
+      "eval_loss": 2.494770050048828,
+      "eval_runtime": 5.1799,
+      "eval_samples_per_second": 96.528,
+      "eval_steps_per_second": 12.162,
+      "step": 1250
+    },
+    {
+      "epoch": 5.0,
+      "eval_exact_match": 17.2,
+      "eval_f1": 25.84752525252526,
+      "step": 1250
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 2.4878485202789307,
+      "learning_rate": 5e-05,
+      "loss": 1.6305,
+      "step": 1300
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 2.599329710006714,
+      "learning_rate": 5e-05,
+      "loss": 1.5477,
+      "step": 1400
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.5542759895324707,
+      "learning_rate": 5e-05,
+      "loss": 1.5708,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5065079365079365,
+      "eval_loss": 2.5797038078308105,
+      "eval_runtime": 5.6827,
+      "eval_samples_per_second": 87.986,
+      "eval_steps_per_second": 11.086,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 17.0,
+      "eval_f1": 25.310858585858586,
+      "step": 1500
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 3.1968846321105957,
+      "learning_rate": 5e-05,
+      "loss": 1.3896,
+      "step": 1600
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 3.8637375831604004,
+      "learning_rate": 5e-05,
+      "loss": 1.421,
+      "step": 1700
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5038412698412699,
+      "eval_loss": 2.6977317333221436,
+      "eval_runtime": 5.3312,
+      "eval_samples_per_second": 93.787,
+      "eval_steps_per_second": 11.817,
+      "step": 1750
+    },
+    {
+      "epoch": 7.0,
+      "eval_exact_match": 16.0,
+      "eval_f1": 23.99441558441559,
+      "step": 1750
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 3.468459367752075,
+      "learning_rate": 5e-05,
+      "loss": 1.3366,
+      "step": 1800
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 3.5727667808532715,
+      "learning_rate": 5e-05,
+      "loss": 1.2585,
+      "step": 1900
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.508258819580078,
+      "learning_rate": 5e-05,
+      "loss": 1.296,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5013333333333333,
+      "eval_loss": 2.786928653717041,
+      "eval_runtime": 5.8398,
+      "eval_samples_per_second": 85.62,
+      "eval_steps_per_second": 10.788,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 15.4,
+      "eval_f1": 22.542871572871572,
+      "step": 2000
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 4.042112350463867,
+      "learning_rate": 5e-05,
+      "loss": 1.1084,
+      "step": 2100
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 3.7548155784606934,
+      "learning_rate": 5e-05,
+      "loss": 1.1457,
+      "step": 2200
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.49692063492063493,
+      "eval_loss": 2.9215197563171387,
+      "eval_runtime": 5.1412,
+      "eval_samples_per_second": 97.254,
+      "eval_steps_per_second": 12.254,
+      "step": 2250
+    },
+    {
+      "epoch": 9.0,
+      "eval_exact_match": 15.0,
+      "eval_f1": 21.59087301587302,
+      "step": 2250
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 4.0875773429870605,
+      "learning_rate": 5e-05,
+      "loss": 1.0677,
+      "step": 2300
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 4.143901348114014,
+      "learning_rate": 5e-05,
+      "loss": 1.0028,
+      "step": 2400
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 5.0514445304870605,
+      "learning_rate": 5e-05,
+      "loss": 1.0401,
+      "step": 2500
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.4953015873015873,
+      "eval_loss": 3.054328203201294,
+      "eval_runtime": 5.4195,
+      "eval_samples_per_second": 92.259,
+      "eval_steps_per_second": 11.625,
+      "step": 2500
+    },
+    {
+      "epoch": 10.0,
+      "eval_exact_match": 15.4,
+      "eval_f1": 21.505119880119878,
+      "step": 2500
+    }
+  ],
+  "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": 6.467692909717094e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-2500/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ede3565a5658e9b8743d6efdd527fd23709623a130469c81f1191d07c33fbcb1
+size 5304