Training in progress, epoch 8, checkpoint

checkpoint-3073/README.md

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

checkpoint-3073/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-3073/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cf85f00b278982cfd473ddf27475f915b6083fa5dd36c96fc16edc83862332a
+size 112130416

checkpoint-3073/added_tokens.json

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

checkpoint-3073/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b644eacc5ac98776e15132df8eaebdccae676ff675a3d2d227dc3dd91a954b4a
+size 224537202

checkpoint-3073/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8f60b350eb068b8df170a9b0555f3135e6ab0a09369e3d16f4bade064529d16
+size 15024

checkpoint-3073/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e735a671ec875e5ef4adcb207df2e1a3df9ea947bc180d7082f3f7db98fc270d
+size 15024

checkpoint-3073/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b145b574a19314e6f08a02f4852d6ef83feed75f0d89e3aff2d6fa616294e70
+size 15024

checkpoint-3073/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aba9a26fbb6e0d8ecacb12208ee15249b962d7133e641f0bd91bf288d0873f9b
+size 15024

checkpoint-3073/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f41765174ab79ac5ac4e4277c9f10736588a44af43d6a0cf83232db434f6ac79
+size 1064

checkpoint-3073/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-3073/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-3073/trainer_state.json

@@ -0,0 +1,387 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 8.998535871156662,
+  "eval_steps": 500,
+  "global_step": 3073,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.29282576866764276,
+      "grad_norm": 0.2425622195005417,
+      "learning_rate": 0.0005,
+      "loss": 1.972,
+      "step": 100
+    },
+    {
+      "epoch": 0.5856515373352855,
+      "grad_norm": 0.24936461448669434,
+      "learning_rate": 0.0005,
+      "loss": 1.954,
+      "step": 200
+    },
+    {
+      "epoch": 0.8784773060029283,
+      "grad_norm": 0.28534355759620667,
+      "learning_rate": 0.0005,
+      "loss": 1.9636,
+      "step": 300
+    },
+    {
+      "epoch": 0.9985358711566618,
+      "eval_accuracy": 0.4142051282051282,
+      "eval_loss": 4.36731481552124,
+      "eval_runtime": 5.2053,
+      "eval_samples_per_second": 96.056,
+      "eval_steps_per_second": 12.103,
+      "step": 341
+    },
+    {
+      "epoch": 0.9985358711566618,
+      "eval_bleu": 0.16271683142618718,
+      "eval_exact_match": 0.0,
+      "eval_prefix_exact_match": 0.032,
+      "step": 341
+    },
+    {
+      "epoch": 1.171303074670571,
+      "grad_norm": 0.39824551343917847,
+      "learning_rate": 0.0005,
+      "loss": 1.8098,
+      "step": 400
+    },
+    {
+      "epoch": 1.4641288433382138,
+      "grad_norm": 0.4451245069503784,
+      "learning_rate": 0.0005,
+      "loss": 1.6717,
+      "step": 500
+    },
+    {
+      "epoch": 1.7569546120058566,
+      "grad_norm": 0.49089929461479187,
+      "learning_rate": 0.0005,
+      "loss": 1.7006,
+      "step": 600
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.43338461538461537,
+      "eval_loss": 4.586850643157959,
+      "eval_runtime": 5.9085,
+      "eval_samples_per_second": 84.624,
+      "eval_steps_per_second": 10.663,
+      "step": 683
+    },
+    {
+      "epoch": 2.0,
+      "eval_bleu": 0.2955390058097942,
+      "eval_exact_match": 0.0,
+      "eval_prefix_exact_match": 0.036,
+      "step": 683
+    },
+    {
+      "epoch": 2.049780380673499,
+      "grad_norm": 0.6150477528572083,
+      "learning_rate": 0.0005,
+      "loss": 1.6383,
+      "step": 700
+    },
+    {
+      "epoch": 2.342606149341142,
+      "grad_norm": 0.5455761551856995,
+      "learning_rate": 0.0005,
+      "loss": 1.2533,
+      "step": 800
+    },
+    {
+      "epoch": 2.6354319180087846,
+      "grad_norm": 0.6252492666244507,
+      "learning_rate": 0.0005,
+      "loss": 1.3256,
+      "step": 900
+    },
+    {
+      "epoch": 2.9282576866764276,
+      "grad_norm": 0.6721358299255371,
+      "learning_rate": 0.0005,
+      "loss": 1.3696,
+      "step": 1000
+    },
+    {
+      "epoch": 2.998535871156662,
+      "eval_accuracy": 0.42538461538461536,
+      "eval_loss": 4.720473766326904,
+      "eval_runtime": 5.6052,
+      "eval_samples_per_second": 89.204,
+      "eval_steps_per_second": 11.24,
+      "step": 1024
+    },
+    {
+      "epoch": 2.998535871156662,
+      "eval_bleu": 0.22572800415324107,
+      "eval_exact_match": 0.004,
+      "eval_prefix_exact_match": 0.044,
+      "step": 1024
+    },
+    {
+      "epoch": 3.22108345534407,
+      "grad_norm": 0.7449648380279541,
+      "learning_rate": 0.0005,
+      "loss": 1.0137,
+      "step": 1100
+    },
+    {
+      "epoch": 3.513909224011713,
+      "grad_norm": 0.7608182430267334,
+      "learning_rate": 0.0005,
+      "loss": 0.9535,
+      "step": 1200
+    },
+    {
+      "epoch": 3.8067349926793557,
+      "grad_norm": 0.8320977687835693,
+      "learning_rate": 0.0005,
+      "loss": 1.0297,
+      "step": 1300
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.41717948717948716,
+      "eval_loss": 4.783276557922363,
+      "eval_runtime": 5.4398,
+      "eval_samples_per_second": 91.915,
+      "eval_steps_per_second": 11.581,
+      "step": 1366
+    },
+    {
+      "epoch": 4.0,
+      "eval_bleu": 0.28487206250718533,
+      "eval_exact_match": 0.008,
+      "eval_prefix_exact_match": 0.034,
+      "step": 1366
+    },
+    {
+      "epoch": 4.099560761346998,
+      "grad_norm": 0.7788870334625244,
+      "learning_rate": 0.0005,
+      "loss": 0.9222,
+      "step": 1400
+    },
+    {
+      "epoch": 4.392386530014641,
+      "grad_norm": 0.7981182336807251,
+      "learning_rate": 0.0005,
+      "loss": 0.6733,
+      "step": 1500
+    },
+    {
+      "epoch": 4.685212298682284,
+      "grad_norm": 0.938538670539856,
+      "learning_rate": 0.0005,
+      "loss": 0.7366,
+      "step": 1600
+    },
+    {
+      "epoch": 4.978038067349927,
+      "grad_norm": 0.9696006178855896,
+      "learning_rate": 0.0005,
+      "loss": 0.7991,
+      "step": 1700
+    },
+    {
+      "epoch": 4.998535871156662,
+      "eval_accuracy": 0.4169230769230769,
+      "eval_loss": 4.9334001541137695,
+      "eval_runtime": 5.1107,
+      "eval_samples_per_second": 97.835,
+      "eval_steps_per_second": 12.327,
+      "step": 1707
+    },
+    {
+      "epoch": 4.998535871156662,
+      "eval_bleu": 0.2463806011037282,
+      "eval_exact_match": 0.006,
+      "eval_prefix_exact_match": 0.034,
+      "step": 1707
+    },
+    {
+      "epoch": 5.270863836017569,
+      "grad_norm": 0.8532861471176147,
+      "learning_rate": 0.0005,
+      "loss": 0.5011,
+      "step": 1800
+    },
+    {
+      "epoch": 5.563689604685212,
+      "grad_norm": 0.8303696513175964,
+      "learning_rate": 0.0005,
+      "loss": 0.5186,
+      "step": 1900
+    },
+    {
+      "epoch": 5.856515373352855,
+      "grad_norm": 1.0248757600784302,
+      "learning_rate": 0.0005,
+      "loss": 0.5801,
+      "step": 2000
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.4133846153846154,
+      "eval_loss": 5.196371078491211,
+      "eval_runtime": 5.382,
+      "eval_samples_per_second": 92.902,
+      "eval_steps_per_second": 11.706,
+      "step": 2049
+    },
+    {
+      "epoch": 6.0,
+      "eval_bleu": 0.24094851152622576,
+      "eval_exact_match": 0.008,
+      "eval_prefix_exact_match": 0.024,
+      "step": 2049
+    },
+    {
+      "epoch": 6.149341142020498,
+      "grad_norm": 0.8996496200561523,
+      "learning_rate": 0.0005,
+      "loss": 0.4772,
+      "step": 2100
+    },
+    {
+      "epoch": 6.44216691068814,
+      "grad_norm": 0.9010413885116577,
+      "learning_rate": 0.0005,
+      "loss": 0.377,
+      "step": 2200
+    },
+    {
+      "epoch": 6.734992679355783,
+      "grad_norm": 0.9476897716522217,
+      "learning_rate": 0.0005,
+      "loss": 0.4228,
+      "step": 2300
+    },
+    {
+      "epoch": 6.998535871156662,
+      "eval_accuracy": 0.4101025641025641,
+      "eval_loss": 5.403963565826416,
+      "eval_runtime": 5.2963,
+      "eval_samples_per_second": 94.406,
+      "eval_steps_per_second": 11.895,
+      "step": 2390
+    },
+    {
+      "epoch": 6.998535871156662,
+      "eval_bleu": 0.20794741740418707,
+      "eval_exact_match": 0.0,
+      "eval_prefix_exact_match": 0.024,
+      "step": 2390
+    },
+    {
+      "epoch": 7.027818448023426,
+      "grad_norm": 0.912433922290802,
+      "learning_rate": 0.0005,
+      "loss": 0.4456,
+      "step": 2400
+    },
+    {
+      "epoch": 7.320644216691068,
+      "grad_norm": 0.8055697679519653,
+      "learning_rate": 0.0005,
+      "loss": 0.2907,
+      "step": 2500
+    },
+    {
+      "epoch": 7.613469985358711,
+      "grad_norm": 0.9217767119407654,
+      "learning_rate": 0.0005,
+      "loss": 0.3285,
+      "step": 2600
+    },
+    {
+      "epoch": 7.906295754026354,
+      "grad_norm": 0.9473926424980164,
+      "learning_rate": 0.0005,
+      "loss": 0.3691,
+      "step": 2700
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.40984615384615386,
+      "eval_loss": 5.655345439910889,
+      "eval_runtime": 5.2073,
+      "eval_samples_per_second": 96.019,
+      "eval_steps_per_second": 12.098,
+      "step": 2732
+    },
+    {
+      "epoch": 8.0,
+      "eval_bleu": 0.1614197770036361,
+      "eval_exact_match": 0.0,
+      "eval_prefix_exact_match": 0.028,
+      "step": 2732
+    },
+    {
+      "epoch": 8.199121522693996,
+      "grad_norm": 0.7216379642486572,
+      "learning_rate": 0.0005,
+      "loss": 0.296,
+      "step": 2800
+    },
+    {
+      "epoch": 8.49194729136164,
+      "grad_norm": 0.8579537868499756,
+      "learning_rate": 0.0005,
+      "loss": 0.2711,
+      "step": 2900
+    },
+    {
+      "epoch": 8.784773060029282,
+      "grad_norm": 0.8787992000579834,
+      "learning_rate": 0.0005,
+      "loss": 0.3052,
+      "step": 3000
+    },
+    {
+      "epoch": 8.998535871156662,
+      "eval_accuracy": 0.41215384615384615,
+      "eval_loss": 5.5592756271362305,
+      "eval_runtime": 5.4045,
+      "eval_samples_per_second": 92.515,
+      "eval_steps_per_second": 11.657,
+      "step": 3073
+    },
+    {
+      "epoch": 8.998535871156662,
+      "eval_bleu": 0.24598687767703112,
+      "eval_exact_match": 0.002,
+      "eval_prefix_exact_match": 0.018,
+      "step": 3073
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 17050,
+  "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": 2.639223295311872e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-3073/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e09109d7fc8f49ffd654f0e1286ef00d8840620753fec453a60a103c55187142
+size 5304