Training in progress, step 14000, checkpoint

checkpoint-14000/config.json

@@ -0,0 +1,39 @@
+{
+  "_name_or_path": "gpt2",
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPT2LMHeadModel"
+  ],
+  "attn_pdrop": 0.1,
+  "bos_token_id": 5,
+  "embd_pdrop": 0.1,
+  "eos_token_id": 4,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gpt2",
+  "n_ctx": 1024,
+  "n_embd": 128,
+  "n_head": 4,
+  "n_inner": null,
+  "n_layer": 4,
+  "n_positions": 128,
+  "reorder_and_upcast_attn": false,
+  "resid_pdrop": 0.1,
+  "scale_attn_by_inverse_layer_idx": false,
+  "scale_attn_weights": true,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "task_specific_params": {
+    "text-generation": {
+      "do_sample": true,
+      "max_length": 50
+    }
+  },
+  "torch_dtype": "float32",
+  "transformers_version": "4.45.2",
+  "use_cache": true,
+  "vocab_size": 8192
+}

checkpoint-14000/generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 5,
+  "eos_token_id": 4,
+  "transformers_version": "4.45.2"
+}

checkpoint-14000/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22cb5b97abe5880a2478be9e68a8764f485acb24c3dd85cb3c851463666ab116
+size 7438256

checkpoint-14000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47f3f88f672c8d2c001fdb34b1171d552768a79d1ec1257aad5557ac4b753035
+size 14910010

checkpoint-14000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b5d9c7665fe797695710125d300d1b8a0afddf84cb697a1b178e5e3d752e5f6
+size 14244

checkpoint-14000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3acd5467204ef3593faf06d029a340fd63a4ed63598a28703adf8adb8251b4c7
+size 1192

checkpoint-14000/special_tokens_map.json

@@ -0,0 +1,17 @@
+{
+  "bos_token": {
+    "content": "[BOS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "[EOS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "[EOS]"
+}

checkpoint-14000/tokenizer_config.json

@@ -0,0 +1,58 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "[EOS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "[BOS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "[BOS]",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "[EOS]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[EOS]",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}

checkpoint-14000/trainer_state.json

@@ -0,0 +1,350 @@
+{
+  "best_metric": 3.499809980392456,
+  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained/reduce_on_plateau_convergence/reverse_transcript_conv/checkpoint-14000",
+  "epoch": 0.3555374965081139,
+  "eval_steps": 1000,
+  "global_step": 14000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.012697767732432638,
+      "grad_norm": 0.7031459212303162,
+      "learning_rate": 0.0001,
+      "loss": 5.9982,
+      "step": 500
+    },
+    {
+      "epoch": 0.025395535464865275,
+      "grad_norm": 0.9951984286308289,
+      "learning_rate": 0.0001,
+      "loss": 4.6781,
+      "step": 1000
+    },
+    {
+      "epoch": 0.025395535464865275,
+      "eval_loss": 4.492164134979248,
+      "eval_runtime": 1.8887,
+      "eval_samples_per_second": 926.554,
+      "eval_steps_per_second": 29.12,
+      "step": 1000
+    },
+    {
+      "epoch": 0.038093303197297916,
+      "grad_norm": 1.1115003824234009,
+      "learning_rate": 0.0001,
+      "loss": 4.4287,
+      "step": 1500
+    },
+    {
+      "epoch": 0.05079107092973055,
+      "grad_norm": 1.2646763324737549,
+      "learning_rate": 0.0001,
+      "loss": 4.3077,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05079107092973055,
+      "eval_loss": 4.193817615509033,
+      "eval_runtime": 1.9023,
+      "eval_samples_per_second": 919.931,
+      "eval_steps_per_second": 28.912,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06348883866216319,
+      "grad_norm": 1.2905216217041016,
+      "learning_rate": 0.0001,
+      "loss": 4.1818,
+      "step": 2500
+    },
+    {
+      "epoch": 0.07618660639459583,
+      "grad_norm": 1.5324379205703735,
+      "learning_rate": 0.0001,
+      "loss": 4.112,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07618660639459583,
+      "eval_loss": 4.043661117553711,
+      "eval_runtime": 1.8938,
+      "eval_samples_per_second": 924.072,
+      "eval_steps_per_second": 29.042,
+      "step": 3000
+    },
+    {
+      "epoch": 0.08888437412702847,
+      "grad_norm": 1.7770308256149292,
+      "learning_rate": 0.0001,
+      "loss": 4.0753,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1015821418594611,
+      "grad_norm": 1.4286599159240723,
+      "learning_rate": 0.0001,
+      "loss": 4.0246,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1015821418594611,
+      "eval_loss": 3.935971260070801,
+      "eval_runtime": 1.8952,
+      "eval_samples_per_second": 923.363,
+      "eval_steps_per_second": 29.02,
+      "step": 4000
+    },
+    {
+      "epoch": 0.11427990959189374,
+      "grad_norm": 1.7079145908355713,
+      "learning_rate": 0.0001,
+      "loss": 3.9216,
+      "step": 4500
+    },
+    {
+      "epoch": 0.12697767732432638,
+      "grad_norm": 1.6703156232833862,
+      "learning_rate": 0.0001,
+      "loss": 3.9453,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12697767732432638,
+      "eval_loss": 3.8490827083587646,
+      "eval_runtime": 1.907,
+      "eval_samples_per_second": 917.653,
+      "eval_steps_per_second": 28.841,
+      "step": 5000
+    },
+    {
+      "epoch": 0.13967544505675902,
+      "grad_norm": 1.6159180402755737,
+      "learning_rate": 0.0001,
+      "loss": 3.8916,
+      "step": 5500
+    },
+    {
+      "epoch": 0.15237321278919166,
+      "grad_norm": 1.5540586709976196,
+      "learning_rate": 0.0001,
+      "loss": 3.8408,
+      "step": 6000
+    },
+    {
+      "epoch": 0.15237321278919166,
+      "eval_loss": 3.8077876567840576,
+      "eval_runtime": 1.9312,
+      "eval_samples_per_second": 906.171,
+      "eval_steps_per_second": 28.48,
+      "step": 6000
+    },
+    {
+      "epoch": 0.1650709805216243,
+      "grad_norm": 1.6115421056747437,
+      "learning_rate": 0.0001,
+      "loss": 3.8216,
+      "step": 6500
+    },
+    {
+      "epoch": 0.17776874825405695,
+      "grad_norm": 1.6438881158828735,
+      "learning_rate": 0.0001,
+      "loss": 3.8155,
+      "step": 7000
+    },
+    {
+      "epoch": 0.17776874825405695,
+      "eval_loss": 3.724748373031616,
+      "eval_runtime": 2.2357,
+      "eval_samples_per_second": 782.742,
+      "eval_steps_per_second": 24.6,
+      "step": 7000
+    },
+    {
+      "epoch": 0.1904665159864896,
+      "grad_norm": 1.5831265449523926,
+      "learning_rate": 0.0001,
+      "loss": 3.7525,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2031642837189222,
+      "grad_norm": 1.6130567789077759,
+      "learning_rate": 0.0001,
+      "loss": 3.7213,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2031642837189222,
+      "eval_loss": 3.696767807006836,
+      "eval_runtime": 1.8961,
+      "eval_samples_per_second": 922.949,
+      "eval_steps_per_second": 29.007,
+      "step": 8000
+    },
+    {
+      "epoch": 0.21586205145135484,
+      "grad_norm": 1.789615273475647,
+      "learning_rate": 0.0001,
+      "loss": 3.7407,
+      "step": 8500
+    },
+    {
+      "epoch": 0.22855981918378748,
+      "grad_norm": 1.7753719091415405,
+      "learning_rate": 0.0001,
+      "loss": 3.7151,
+      "step": 9000
+    },
+    {
+      "epoch": 0.22855981918378748,
+      "eval_loss": 3.6513242721557617,
+      "eval_runtime": 1.8933,
+      "eval_samples_per_second": 924.291,
+      "eval_steps_per_second": 29.049,
+      "step": 9000
+    },
+    {
+      "epoch": 0.24125758691622012,
+      "grad_norm": 1.5469183921813965,
+      "learning_rate": 0.0001,
+      "loss": 3.6057,
+      "step": 9500
+    },
+    {
+      "epoch": 0.25395535464865276,
+      "grad_norm": 1.7085388898849487,
+      "learning_rate": 0.0001,
+      "loss": 3.7075,
+      "step": 10000
+    },
+    {
+      "epoch": 0.25395535464865276,
+      "eval_loss": 3.600736141204834,
+      "eval_runtime": 1.8818,
+      "eval_samples_per_second": 929.961,
+      "eval_steps_per_second": 29.227,
+      "step": 10000
+    },
+    {
+      "epoch": 0.2666531223810854,
+      "grad_norm": 1.5484737157821655,
+      "learning_rate": 0.0001,
+      "loss": 3.6486,
+      "step": 10500
+    },
+    {
+      "epoch": 0.27935089011351805,
+      "grad_norm": 2.095698118209839,
+      "learning_rate": 0.0001,
+      "loss": 3.5585,
+      "step": 11000
+    },
+    {
+      "epoch": 0.27935089011351805,
+      "eval_loss": 3.584713935852051,
+      "eval_runtime": 1.901,
+      "eval_samples_per_second": 920.585,
+      "eval_steps_per_second": 28.933,
+      "step": 11000
+    },
+    {
+      "epoch": 0.2920486578459507,
+      "grad_norm": 1.822464942932129,
+      "learning_rate": 0.0001,
+      "loss": 3.6714,
+      "step": 11500
+    },
+    {
+      "epoch": 0.30474642557838333,
+      "grad_norm": 1.756412148475647,
+      "learning_rate": 0.0001,
+      "loss": 3.6149,
+      "step": 12000
+    },
+    {
+      "epoch": 0.30474642557838333,
+      "eval_loss": 3.5466842651367188,
+      "eval_runtime": 1.8965,
+      "eval_samples_per_second": 922.772,
+      "eval_steps_per_second": 29.001,
+      "step": 12000
+    },
+    {
+      "epoch": 0.31744419331081597,
+      "grad_norm": 1.7527214288711548,
+      "learning_rate": 0.0001,
+      "loss": 3.4993,
+      "step": 12500
+    },
+    {
+      "epoch": 0.3301419610432486,
+      "grad_norm": 1.4713279008865356,
+      "learning_rate": 0.0001,
+      "loss": 3.5912,
+      "step": 13000
+    },
+    {
+      "epoch": 0.3301419610432486,
+      "eval_loss": 3.5182790756225586,
+      "eval_runtime": 1.8952,
+      "eval_samples_per_second": 923.389,
+      "eval_steps_per_second": 29.021,
+      "step": 13000
+    },
+    {
+      "epoch": 0.34283972877568125,
+      "grad_norm": 1.8965306282043457,
+      "learning_rate": 0.0001,
+      "loss": 3.5699,
+      "step": 13500
+    },
+    {
+      "epoch": 0.3555374965081139,
+      "grad_norm": 1.6826404333114624,
+      "learning_rate": 0.0001,
+      "loss": 3.4807,
+      "step": 14000
+    },
+    {
+      "epoch": 0.3555374965081139,
+      "eval_loss": 3.499809980392456,
+      "eval_runtime": 1.922,
+      "eval_samples_per_second": 910.515,
+      "eval_steps_per_second": 28.616,
+      "step": 14000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 39377,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.001
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 272961110016000.0,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-14000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:539b27162b2f1812097821a7135b4ffe9aa17afeac56bc6323495680f90e1a25
+size 5432