Training in progress, step 11000, checkpoint

checkpoint-11000/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-11000/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-11000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e903c8192088cc20bfbad6524ded36e628b87bdc9a0177b9a612910063e0a56
+size 272123144

checkpoint-11000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:627d937aba04e2b25fb1ba6d976248bf757c30ca2c801e00d88fede6e10bfdb7
+size 544259743

checkpoint-11000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d3b7102895eb0637b0cab516bd672f216b2bf79078a83eb301011a90444f44c
+size 14244

checkpoint-11000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1375fab72e2777b05d344b00d5eccb644359746eaa8ac3c71ac8f89aa6612828
+size 1000

checkpoint-11000/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-11000/tokenizer.model

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

checkpoint-11000/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-11000/trainer_state.json

@@ -0,0 +1,286 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.908514013749339,
+  "eval_steps": 1000,
+  "global_step": 11000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.13220518244315177,
+      "grad_norm": 0.8546391725540161,
+      "learning_rate": 8.816009873931059e-05,
+      "loss": 5.1118,
+      "step": 500
+    },
+    {
+      "epoch": 0.26441036488630354,
+      "grad_norm": 0.8593688607215881,
+      "learning_rate": 9.59831475011252e-05,
+      "loss": 3.406,
+      "step": 1000
+    },
+    {
+      "epoch": 0.26441036488630354,
+      "eval_accuracy": 0.5035306174465283,
+      "eval_loss": 3.23445987701416,
+      "eval_runtime": 74.0522,
+      "eval_samples_per_second": 24.847,
+      "eval_steps_per_second": 0.527,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3966155473294553,
+      "grad_norm": 0.9617258906364441,
+      "learning_rate": 9.134314230431938e-05,
+      "loss": 3.0005,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5288207297726071,
+      "grad_norm": 0.8953185677528381,
+      "learning_rate": 8.670313710751356e-05,
+      "loss": 2.8119,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5288207297726071,
+      "eval_accuracy": 0.5365118094348038,
+      "eval_loss": 2.821384906768799,
+      "eval_runtime": 72.8065,
+      "eval_samples_per_second": 25.272,
+      "eval_steps_per_second": 0.536,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6610259122157589,
+      "grad_norm": 1.4154396057128906,
+      "learning_rate": 8.206313191070773e-05,
+      "loss": 2.686,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7932310946589106,
+      "grad_norm": 1.821349024772644,
+      "learning_rate": 7.742312671390191e-05,
+      "loss": 2.607,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7932310946589106,
+      "eval_accuracy": 0.5497897240925214,
+      "eval_loss": 2.657219886779785,
+      "eval_runtime": 72.9553,
+      "eval_samples_per_second": 25.221,
+      "eval_steps_per_second": 0.535,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9254362771020624,
+      "grad_norm": 2.0297396183013916,
+      "learning_rate": 7.278312151709609e-05,
+      "loss": 2.5642,
+      "step": 3500
+    },
+    {
+      "epoch": 1.0576414595452142,
+      "grad_norm": 2.8318285942077637,
+      "learning_rate": 6.814311632029027e-05,
+      "loss": 2.4734,
+      "step": 4000
+    },
+    {
+      "epoch": 1.0576414595452142,
+      "eval_accuracy": 0.5582058048894458,
+      "eval_loss": 2.5735702514648438,
+      "eval_runtime": 73.5253,
+      "eval_samples_per_second": 25.025,
+      "eval_steps_per_second": 0.53,
+      "step": 4000
+    },
+    {
+      "epoch": 1.189846641988366,
+      "grad_norm": 2.178349018096924,
+      "learning_rate": 6.350311112348446e-05,
+      "loss": 2.4542,
+      "step": 4500
+    },
+    {
+      "epoch": 1.3220518244315178,
+      "grad_norm": 2.944026470184326,
+      "learning_rate": 5.886310592667864e-05,
+      "loss": 2.4335,
+      "step": 5000
+    },
+    {
+      "epoch": 1.3220518244315178,
+      "eval_accuracy": 0.5617462458316518,
+      "eval_loss": 2.5335800647735596,
+      "eval_runtime": 73.7726,
+      "eval_samples_per_second": 24.941,
+      "eval_steps_per_second": 0.529,
+      "step": 5000
+    },
+    {
+      "epoch": 1.4542570068746694,
+      "grad_norm": 3.0737102031707764,
+      "learning_rate": 5.422310072987282e-05,
+      "loss": 2.4012,
+      "step": 5500
+    },
+    {
+      "epoch": 1.5864621893178212,
+      "grad_norm": 3.8758912086486816,
+      "learning_rate": 4.9583095533066995e-05,
+      "loss": 2.3858,
+      "step": 6000
+    },
+    {
+      "epoch": 1.5864621893178212,
+      "eval_accuracy": 0.5662889488328625,
+      "eval_loss": 2.4896483421325684,
+      "eval_runtime": 73.8595,
+      "eval_samples_per_second": 24.912,
+      "eval_steps_per_second": 0.528,
+      "step": 6000
+    },
+    {
+      "epoch": 1.718667371760973,
+      "grad_norm": 7.714890956878662,
+      "learning_rate": 4.4943090336261176e-05,
+      "loss": 2.3797,
+      "step": 6500
+    },
+    {
+      "epoch": 1.8508725542041247,
+      "grad_norm": 7.321514129638672,
+      "learning_rate": 4.030308513945535e-05,
+      "loss": 2.374,
+      "step": 7000
+    },
+    {
+      "epoch": 1.8508725542041247,
+      "eval_accuracy": 0.5680874450415242,
+      "eval_loss": 2.4711225032806396,
+      "eval_runtime": 73.2926,
+      "eval_samples_per_second": 25.105,
+      "eval_steps_per_second": 0.532,
+      "step": 7000
+    },
+    {
+      "epoch": 1.9830777366472767,
+      "grad_norm": 5.710904598236084,
+      "learning_rate": 3.566307994264953e-05,
+      "loss": 2.3445,
+      "step": 7500
+    },
+    {
+      "epoch": 2.1152829190904283,
+      "grad_norm": 16.555458068847656,
+      "learning_rate": 3.1023074745843715e-05,
+      "loss": 2.3044,
+      "step": 8000
+    },
+    {
+      "epoch": 2.1152829190904283,
+      "eval_accuracy": 0.5701761326225017,
+      "eval_loss": 2.4485366344451904,
+      "eval_runtime": 73.4687,
+      "eval_samples_per_second": 25.045,
+      "eval_steps_per_second": 0.531,
+      "step": 8000
+    },
+    {
+      "epoch": 2.24748810153358,
+      "grad_norm": 6.3735737800598145,
+      "learning_rate": 2.6383069549037897e-05,
+      "loss": 2.2922,
+      "step": 8500
+    },
+    {
+      "epoch": 2.379693283976732,
+      "grad_norm": 19.310932159423828,
+      "learning_rate": 2.1743064352232075e-05,
+      "loss": 2.2873,
+      "step": 9000
+    },
+    {
+      "epoch": 2.379693283976732,
+      "eval_accuracy": 0.5747236146216096,
+      "eval_loss": 2.413153886795044,
+      "eval_runtime": 73.9,
+      "eval_samples_per_second": 24.899,
+      "eval_steps_per_second": 0.528,
+      "step": 9000
+    },
+    {
+      "epoch": 2.5118984664198836,
+      "grad_norm": 6.743342876434326,
+      "learning_rate": 1.7103059155426253e-05,
+      "loss": 2.2626,
+      "step": 9500
+    },
+    {
+      "epoch": 2.6441036488630356,
+      "grad_norm": 9.461324691772461,
+      "learning_rate": 1.2463053958620433e-05,
+      "loss": 2.2669,
+      "step": 10000
+    },
+    {
+      "epoch": 2.6441036488630356,
+      "eval_accuracy": 0.5778246532571526,
+      "eval_loss": 2.387904167175293,
+      "eval_runtime": 73.5959,
+      "eval_samples_per_second": 25.001,
+      "eval_steps_per_second": 0.53,
+      "step": 10000
+    },
+    {
+      "epoch": 2.776308831306187,
+      "grad_norm": 10.6486234664917,
+      "learning_rate": 7.823048761814613e-06,
+      "loss": 2.2557,
+      "step": 10500
+    },
+    {
+      "epoch": 2.908514013749339,
+      "grad_norm": 8.470663070678711,
+      "learning_rate": 3.183043565008793e-06,
+      "loss": 2.2463,
+      "step": 11000
+    },
+    {
+      "epoch": 2.908514013749339,
+      "eval_accuracy": 0.5807555064675771,
+      "eval_loss": 2.363588571548462,
+      "eval_runtime": 73.5703,
+      "eval_samples_per_second": 25.01,
+      "eval_steps_per_second": 0.53,
+      "step": 11000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 11346,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4094227888249242e+17,
+  "train_batch_size": 24,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-11000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84a8c3b8deb94d253a20c7d637102edeaa129f75276f850abb181ddfd4b1ddcf
+size 5176