Training in progress, step 10000, checkpoint

checkpoint-10000/config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "JackFram/llama-68m",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "max_position_embeddings": 2048,
+  "model_type": "llama",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 2,
+  "num_key_value_heads": 12,
+  "pad_token_id": 1,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.41.0.dev0",
+  "use_cache": true,
+  "vocab_size": 32000
+}

checkpoint-10000/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 2,
+  "pad_token_id": 1,
+  "transformers_version": "4.41.0.dev0"
+}

checkpoint-10000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd08bc2bc8a7b7220af43eb52ef4280448bb97bea31947593fb17094fd789061
+size 272123144

checkpoint-10000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f5096577241fd2d09725c49fc82dd6429303d9e0e4b839d163aae2069f5c24c
+size 544259743

checkpoint-10000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3e5d946241df2516b06d7074d8779088eae7607173ad780df56583910a9589b
+size 14244

checkpoint-10000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5366e3edaa59533c3cf8562be8a705d3bdeaeea0f705fb724dd7007bb38e71a9
+size 1000

checkpoint-10000/special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-10000/tokenizer.model

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

checkpoint-10000/tokenizer_config.json

@@ -0,0 +1,45 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": true,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 2048,
+  "pad_token": "<unk>",
+  "padding": "max_length",
+  "return_tensors": "pt",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false,
+  "use_fast": true
+}

checkpoint-10000/trainer_state.json

@@ -0,0 +1,263 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.322226629644321,
+  "eval_steps": 1000,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.06611133148221605,
+      "grad_norm": 1.0232534408569336,
+      "learning_rate": 4.4080049369655294e-05,
+      "loss": 5.6838,
+      "step": 500
+    },
+    {
+      "epoch": 0.1322226629644321,
+      "grad_norm": 1.0166140794754028,
+      "learning_rate": 8.816009873931059e-05,
+      "loss": 3.8378,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1322226629644321,
+      "eval_accuracy": 0.4720710052887577,
+      "eval_loss": 3.6431853771209717,
+      "eval_runtime": 65.5904,
+      "eval_samples_per_second": 28.053,
+      "eval_steps_per_second": 1.174,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19833399444664815,
+      "grad_norm": 1.0467888116836548,
+      "learning_rate": 9.830315009952811e-05,
+      "loss": 3.3712,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2644453259288642,
+      "grad_norm": 1.1263777017593384,
+      "learning_rate": 9.59831475011252e-05,
+      "loss": 3.0922,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2644453259288642,
+      "eval_accuracy": 0.5138609524011809,
+      "eval_loss": 3.076597213745117,
+      "eval_runtime": 64.6228,
+      "eval_samples_per_second": 28.473,
+      "eval_steps_per_second": 1.192,
+      "step": 2000
+    },
+    {
+      "epoch": 0.33055665741108026,
+      "grad_norm": 1.4438892602920532,
+      "learning_rate": 9.366314490272228e-05,
+      "loss": 2.9066,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3966679888932963,
+      "grad_norm": 1.3693314790725708,
+      "learning_rate": 9.134314230431938e-05,
+      "loss": 2.7993,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3966679888932963,
+      "eval_accuracy": 0.5319845054268176,
+      "eval_loss": 2.84745454788208,
+      "eval_runtime": 64.9029,
+      "eval_samples_per_second": 28.35,
+      "eval_steps_per_second": 1.186,
+      "step": 3000
+    },
+    {
+      "epoch": 0.46277932037551234,
+      "grad_norm": 1.3279718160629272,
+      "learning_rate": 8.902313970591646e-05,
+      "loss": 2.7166,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5288906518577284,
+      "grad_norm": 1.465155839920044,
+      "learning_rate": 8.670313710751356e-05,
+      "loss": 2.7115,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5288906518577284,
+      "eval_accuracy": 0.5392130052462777,
+      "eval_loss": 2.7528512477874756,
+      "eval_runtime": 65.039,
+      "eval_samples_per_second": 28.291,
+      "eval_steps_per_second": 1.184,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5950019833399445,
+      "grad_norm": 2.4618444442749023,
+      "learning_rate": 8.438313450911065e-05,
+      "loss": 2.644,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6611133148221605,
+      "grad_norm": 3.049086093902588,
+      "learning_rate": 8.206313191070773e-05,
+      "loss": 2.6702,
+      "step": 5000
+    },
+    {
+      "epoch": 0.6611133148221605,
+      "eval_accuracy": 0.5420291625071685,
+      "eval_loss": 2.7150135040283203,
+      "eval_runtime": 64.9223,
+      "eval_samples_per_second": 28.342,
+      "eval_steps_per_second": 1.186,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7272246463043766,
+      "grad_norm": 3.927698850631714,
+      "learning_rate": 7.974312931230483e-05,
+      "loss": 2.6029,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7933359777865926,
+      "grad_norm": 4.909026622772217,
+      "learning_rate": 7.742312671390191e-05,
+      "loss": 2.6484,
+      "step": 6000
+    },
+    {
+      "epoch": 0.7933359777865926,
+      "eval_accuracy": 0.543187538497483,
+      "eval_loss": 2.696218729019165,
+      "eval_runtime": 64.8633,
+      "eval_samples_per_second": 28.367,
+      "eval_steps_per_second": 1.187,
+      "step": 6000
+    },
+    {
+      "epoch": 0.8594473092688086,
+      "grad_norm": 10.72818660736084,
+      "learning_rate": 7.510312411549901e-05,
+      "loss": 2.6474,
+      "step": 6500
+    },
+    {
+      "epoch": 0.9255586407510247,
+      "grad_norm": 12.435935020446777,
+      "learning_rate": 7.278312151709609e-05,
+      "loss": 2.6419,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9255586407510247,
+      "eval_accuracy": 0.5387701514411334,
+      "eval_loss": 2.7223353385925293,
+      "eval_runtime": 68.3123,
+      "eval_samples_per_second": 26.935,
+      "eval_steps_per_second": 1.127,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9916699722332408,
+      "grad_norm": 15.605013847351074,
+      "learning_rate": 7.046311891869319e-05,
+      "loss": 2.6239,
+      "step": 7500
+    },
+    {
+      "epoch": 1.0577813037154569,
+      "grad_norm": 55.199256896972656,
+      "learning_rate": 6.814311632029027e-05,
+      "loss": 2.5853,
+      "step": 8000
+    },
+    {
+      "epoch": 1.0577813037154569,
+      "eval_accuracy": 0.5401743803232727,
+      "eval_loss": 2.7088677883148193,
+      "eval_runtime": 66.8855,
+      "eval_samples_per_second": 27.51,
+      "eval_steps_per_second": 1.151,
+      "step": 8000
+    },
+    {
+      "epoch": 1.1238926351976728,
+      "grad_norm": 19.066770553588867,
+      "learning_rate": 6.582311372188736e-05,
+      "loss": 2.616,
+      "step": 8500
+    },
+    {
+      "epoch": 1.190003966679889,
+      "grad_norm": 25.54907989501953,
+      "learning_rate": 6.350311112348446e-05,
+      "loss": 2.6009,
+      "step": 9000
+    },
+    {
+      "epoch": 1.190003966679889,
+      "eval_accuracy": 0.5401903103162634,
+      "eval_loss": 2.703549861907959,
+      "eval_runtime": 65.0156,
+      "eval_samples_per_second": 28.301,
+      "eval_steps_per_second": 1.184,
+      "step": 9000
+    },
+    {
+      "epoch": 1.256115298162105,
+      "grad_norm": 24.64689826965332,
+      "learning_rate": 6.118310852508154e-05,
+      "loss": 2.622,
+      "step": 9500
+    },
+    {
+      "epoch": 1.322226629644321,
+      "grad_norm": 39.75895309448242,
+      "learning_rate": 5.886310592667864e-05,
+      "loss": 2.6347,
+      "step": 10000
+    },
+    {
+      "epoch": 1.322226629644321,
+      "eval_accuracy": 0.5368293472950872,
+      "eval_loss": 2.7321841716766357,
+      "eval_runtime": 70.4109,
+      "eval_samples_per_second": 26.132,
+      "eval_steps_per_second": 1.094,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 22689,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 10000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.406996333009306e+16,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-10000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b1788df2e878fccca08354d00fe3ff798a83f0940758c72244a04260d9cb1f0
+size 5176