Upload 8 files

nanoDist18/config.json

@@ -0,0 +1,26 @@
+{
+  "architectures": [
+    "RobertaForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 156,
+  "model_type": "roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.24.0",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 26
+}

nanoDist18/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a767a1603187aee67d00f21574973622e0305ea38ed47df49be4456a18ee749f
+size 686428357

nanoDist18/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad579a984de9c19aef896f0bd3770834b84023d0412aee795471ab0ceeffd341
+size 343225593

nanoDist18/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8161de00b7e4b8b84bf4e81918ab41457046704246e186132ef5b25525c1e979
+size 15597

nanoDist18/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:987e725c68fbad78a31692df7a404121c4a7d55370f9c0db127b90fa8c7c19b3
+size 557

nanoDist18/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28807a4920832c08a4fe739fbabad8636da10fd1f39875f013f6d90a3e33e21d
+size 627

nanoDist18/trainer_state.json

@@ -0,0 +1,58 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "global_step": 139704,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.359719978459882e-05,
+      "loss": 0.6809,
+      "step": 46568
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.3213633894920349,
+      "eval_runtime": 420.6563,
+      "eval_samples_per_second": 2182.138,
+      "eval_steps_per_second": 11.366,
+      "step": 46568
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 6.270866989768444e-05,
+      "loss": 0.4706,
+      "step": 93136
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.2747191786766052,
+      "eval_runtime": 418.1806,
+      "eval_samples_per_second": 2195.056,
+      "eval_steps_per_second": 11.433,
+      "step": 93136
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.1818793753365646e-05,
+      "loss": 0.4215,
+      "step": 139704
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.2559063732624054,
+      "eval_runtime": 418.7155,
+      "eval_samples_per_second": 2192.252,
+      "eval_steps_per_second": 11.418,
+      "step": 139704
+    }
+  ],
+  "max_steps": 232840,
+  "num_train_epochs": 5,
+  "total_flos": 2.1227226187002455e+18,
+  "trial_name": null,
+  "trial_params": null
+}

nanoDist18/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49da3e89dac24b684080ef6f785d204e946deaf066e378059d8d4eb8b0a67934
+size 3387