Training in progress, epoch 2, checkpoint

checkpoint-375/adapter_config.json

@@ -1,6 +1,6 @@
 {
   "auto_mapping": null,
-  "base_model_name_or_path": "Qwen/Qwen1.5-4B",
+  "base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
   "bias": "none",
   "fan_in_fan_out": false,
   "inference_mode": true,

checkpoint-375/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eee0a1d04d185d71b274ac663e7d9bff08199610e0c52b922c743ad251b64bd0
+size 143269386

checkpoint-375/added_tokens.json

@@ -1,5 +1,5 @@
 {
-  "<|endoftext|>": 151643,
-  "<|im_end|>": 151645,
-  "<|im_start|>": 151644
+  "</s>": 2,
+  "<s>": 1,
+  "<unk>": 0
 }

checkpoint-375/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ba181f70c9e17ae519fab3ecd8222e32ed5491bec398dbcb64eca2c4dedbecc
-size 224537202
+oid sha256:b8eaccd444f867e3581d63b002d35ab1fdb10b5d0e688c6cf45d55947cabc436
+size 286585234

checkpoint-375/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ffe0452198a1775d0b1a3be4eb867c6abd6124bbb5457137601b99a292ab513
+oid sha256:e25f60b6a2aedad76c8e0e0320dbef6efe914d871d8ee84d16d8972f0eca0949
 size 15024

checkpoint-375/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d18156f60ff9496f99d6e8597e2804556a6f961cdb63da148bfe79f0adea77e3
+oid sha256:3ef72b5f01911322eabcce5980ccae5f7294320fbb198cbfda8fafbcf2232505
 size 15024

checkpoint-375/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4dfcf6d49a39b317c7ee77599884c4bc614792c14d8c393cd27013ccb5871d02
+oid sha256:9332c9182a574d85ff435ce6cfe8a54edf04db9242321ac4961f00f6f40e511a
 size 15024

checkpoint-375/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6dfdefc8565bff7ff9ce3cd55b31cd1c475ceb10ead1886f615c4c0e60cfcdf7
+oid sha256:378da0be37d1c6f5098ff249d769f1474358c25c69cad26e5e1ce3ecf37cb115
 size 15024

checkpoint-375/special_tokens_map.json

@@ -1,14 +1,6 @@
 {
-  "additional_special_tokens": [
-    "<|im_start|>",
-    "<|im_end|>"
-  ],
-  "eos_token": {
-    "content": "<|endoftext|>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "pad_token": "<|endoftext|>"
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "</s>",
+  "unk_token": "<unk>"
 }

checkpoint-375/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

checkpoint-375/tokenizer_config.json

@@ -1,24 +1,23 @@
 {
-  "add_prefix_space": false,
   "added_tokens_decoder": {
-    "151643": {
-      "content": "<|endoftext|>",
+    "0": {
+      "content": "<unk>",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
       "single_word": false,
       "special": true
     },
-    "151644": {
-      "content": "<|im_start|>",
+    "1": {
+      "content": "<s>",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
       "single_word": false,
       "special": true
     },
-    "151645": {
-      "content": "<|im_end|>",
+    "2": {
+      "content": "</s>",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -26,18 +25,16 @@
       "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 %}",
+  "additional_special_tokens": [],
+  "bos_token": "<s>",
   "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
+  "eos_token": "</s>",
+  "legacy": false,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "padding_side": "left",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": true
 }

checkpoint-375/trainer_state.json

@@ -9,84 +9,67 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.5333333333333333,
-      "grad_norm": 0.2081739604473114,
+      "epoch": 0.53,
       "learning_rate": 0.0005,
-      "loss": 1.7461,
+      "loss": 1.2462,
       "step": 100
     },
     {
-      "epoch": 0.9973333333333333,
-      "eval_accuracy": 0.6093004484304932,
-      "eval_loss": 1.6695144176483154,
-      "eval_runtime": 6.0389,
-      "eval_samples_per_second": 82.796,
-      "eval_steps_per_second": 10.432,
+      "epoch": 1.0,
+      "eval_accuracy": 0.6696470588235294,
+      "eval_loss": 1.1816030740737915,
+      "eval_runtime": 7.5359,
+      "eval_samples_per_second": 66.349,
+      "eval_steps_per_second": 8.36,
       "step": 187
     },
     {
-      "epoch": 0.9973333333333333,
-      "eval_exact_match": 15.8,
-      "eval_f1": 24.64418803418805,
-      "eval_qa_bleu": 12.64258947770803,
-      "eval_qa_exact_match": 0.116,
-      "eval_recite_bleu": 12.676034356605994,
+      "epoch": 1.0,
+      "eval_exact_match": 28.8,
+      "eval_f1": 38.19841269841268,
+      "eval_qa_bleu": 24.052327144564227,
+      "eval_qa_exact_match": 0.216,
+      "eval_recite_bleu": 16.828353493471358,
       "eval_recite_exact_match": 0.002,
       "step": 187
     },
     {
-      "epoch": 1.0666666666666667,
-      "grad_norm": 0.2944607138633728,
+      "epoch": 1.07,
       "learning_rate": 0.0005,
-      "loss": 1.6853,
+      "loss": 1.1798,
       "step": 200
     },
     {
       "epoch": 1.6,
-      "grad_norm": 0.40688613057136536,
       "learning_rate": 0.0005,
-      "loss": 1.4286,
+      "loss": 0.9978,
       "step": 300
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.6086547085201793,
-      "eval_loss": 1.6994106769561768,
-      "eval_runtime": 6.3284,
-      "eval_samples_per_second": 79.009,
-      "eval_steps_per_second": 9.955,
+      "eval_accuracy": 0.6694823529411764,
+      "eval_loss": 1.196727991104126,
+      "eval_runtime": 6.6634,
+      "eval_samples_per_second": 75.037,
+      "eval_steps_per_second": 9.455,
       "step": 375
     },
     {
       "epoch": 2.0,
-      "eval_exact_match": 13.6,
-      "eval_f1": 22.022914862914874,
-      "eval_qa_bleu": 11.191170451488254,
-      "eval_qa_exact_match": 0.098,
-      "eval_recite_bleu": 13.062440471503999,
+      "eval_exact_match": 26.4,
+      "eval_f1": 35.01285714285714,
+      "eval_qa_bleu": 22.554598366442242,
+      "eval_qa_exact_match": 0.202,
+      "eval_recite_bleu": 16.66380860754827,
       "eval_recite_exact_match": 0.008,
       "step": 375
     }
   ],
   "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": 3.777309007518106e+16,
-  "train_batch_size": 1,
+  "total_flos": 8.689846617689293e+16,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-375/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7b6061b8d6135a3f759c6b956a7afa194ea4a0f719fc2e4eb293c4d6c15c2590
-size 5304
+oid sha256:a28079dc7535e79a5e324b19b9af4575c5cbbdcf35a6916d09f5d38aff3423a1
+size 4728