Upload 13 files

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[MASK]": 128000
+}

config.json

@@ -0,0 +1,49 @@
+{
+  "_name_or_path": "microsoft/deberta-v3-xsmall",
+  "architectures": [
+    "DebertaV2ForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 384,
+  "id2label": {
+    "0": "casual",
+    "1": "possibly_needs_caution",
+    "2": "probably_needs_caution",
+    "3": "needs_caution",
+    "4": "needs_intervention"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 1536,
+  "label2id": {
+    "casual": 0,
+    "needs_caution": 3,
+    "needs_intervention": 4,
+    "possibly_needs_caution": 1,
+    "probably_needs_caution": 2
+  },
+  "layer_norm_eps": 1e-07,
+  "max_position_embeddings": 512,
+  "max_relative_positions": -1,
+  "model_type": "deberta-v2",
+  "norm_rel_ebd": "layer_norm",
+  "num_attention_heads": 6,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "pooler_hidden_size": 384,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "position_biased_input": false,
+  "position_buckets": 256,
+  "relative_attention": true,
+  "share_att_key": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.20.1",
+  "type_vocab_size": 0,
+  "vocab_size": 128100
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77773ad0530b499711769b2d5fc366b92b4cac09c54e4b6830ab82b68f92945f
+size 566772057

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d36e58a1c568f2c1e6ac4699a2a7d4cc00cd34cebffecd702f5ee189daeb2c69
+size 283399723

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:995cada29c27c87dddbd3bc2d632974ad965bacba1ca785dd403f97bf45f6831
+size 14503

scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52e2dd89a3ca7aed864a6c463d379b57a2d1774071373c07789c8886ce828b76
+size 559

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aec80074eb6daf203aeac4a2ebafa8af04fdb64c4a549bba3361d3fb3bf8e48d
+size 623

special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
+size 2464616

tokenizer_config.json

@@ -0,0 +1,18 @@
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "max_length": 512,
+  "model_max_length": 512,
+  "name_or_path": "microsoft/deberta-v3-xsmall",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "sp_model_kwargs": {},
+  "special_tokens_map_file": null,
+  "split_by_punct": false,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
+}

trainer_state.json

@@ -0,0 +1,256 @@
+{
+  "best_metric": 1.495205044746399,
+  "best_model_checkpoint": "safety_instruct_detect/checkpoint-2556",
+  "epoch": 10.0,
+  "global_step": 12780,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.2455e-05,
+      "loss": 1.5162,
+      "step": 500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.4955000000000006e-05,
+      "loss": 1.4983,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.4036969420356002,
+      "eval_loss": 1.4984349012374878,
+      "eval_runtime": 11.2442,
+      "eval_samples_per_second": 389.712,
+      "eval_steps_per_second": 48.736,
+      "step": 1278
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.309380305602717e-05,
+      "loss": 1.4868,
+      "step": 1500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.1183786078098476e-05,
+      "loss": 1.4851,
+      "step": 2000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.927376910016978e-05,
+      "loss": 1.4886,
+      "step": 2500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.4036969420356002,
+      "eval_loss": 1.495205044746399,
+      "eval_runtime": 11.4186,
+      "eval_samples_per_second": 383.76,
+      "eval_steps_per_second": 47.992,
+      "step": 2556
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 3.736757215619695e-05,
+      "loss": 1.491,
+      "step": 3000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.545755517826825e-05,
+      "loss": 1.4828,
+      "step": 3500
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.4036969420356002,
+      "eval_loss": 1.4983419179916382,
+      "eval_runtime": 11.4982,
+      "eval_samples_per_second": 381.103,
+      "eval_steps_per_second": 47.66,
+      "step": 3834
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 3.354753820033956e-05,
+      "loss": 1.4937,
+      "step": 4000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.163752122241087e-05,
+      "loss": 1.4751,
+      "step": 4500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.9731324278438033e-05,
+      "loss": 1.4853,
+      "step": 5000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.4036969420356002,
+      "eval_loss": 1.5341633558273315,
+      "eval_runtime": 11.4113,
+      "eval_samples_per_second": 384.006,
+      "eval_steps_per_second": 48.023,
+      "step": 5112
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.782130730050934e-05,
+      "loss": 1.4528,
+      "step": 5500
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.5911290322580646e-05,
+      "loss": 1.4446,
+      "step": 6000
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.39639434048379735,
+      "eval_loss": 1.557020664215088,
+      "eval_runtime": 11.9866,
+      "eval_samples_per_second": 365.575,
+      "eval_steps_per_second": 45.718,
+      "step": 6390
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 2.4001273344651956e-05,
+      "loss": 1.4092,
+      "step": 6500
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.209125636672326e-05,
+      "loss": 1.3309,
+      "step": 7000
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.018123938879457e-05,
+      "loss": 1.3054,
+      "step": 7500
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.36376083979917845,
+      "eval_loss": 1.8100188970565796,
+      "eval_runtime": 11.9522,
+      "eval_samples_per_second": 366.627,
+      "eval_steps_per_second": 45.849,
+      "step": 7668
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1.8271222410865876e-05,
+      "loss": 1.2194,
+      "step": 8000
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1.6361205432937183e-05,
+      "loss": 1.1677,
+      "step": 8500
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.3265632131446828,
+      "eval_loss": 2.0681562423706055,
+      "eval_runtime": 12.1156,
+      "eval_samples_per_second": 361.683,
+      "eval_steps_per_second": 45.231,
+      "step": 8946
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 1.4455008488964348e-05,
+      "loss": 1.1879,
+      "step": 9000
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 1.2548811544991511e-05,
+      "loss": 1.0508,
+      "step": 9500
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 1.063879456706282e-05,
+      "loss": 1.0861,
+      "step": 10000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.3014605203103606,
+      "eval_loss": 2.157534599304199,
+      "eval_runtime": 11.8194,
+      "eval_samples_per_second": 370.747,
+      "eval_steps_per_second": 46.365,
+      "step": 10224
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 8.728777589134126e-06,
+      "loss": 1.0225,
+      "step": 10500
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 6.818760611205433e-06,
+      "loss": 0.9752,
+      "step": 11000
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.912563667232598e-06,
+      "loss": 0.9729,
+      "step": 11500
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.30785029666818803,
+      "eval_loss": 2.388486385345459,
+      "eval_runtime": 11.8239,
+      "eval_samples_per_second": 370.605,
+      "eval_steps_per_second": 46.347,
+      "step": 11502
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 3.002546689303905e-06,
+      "loss": 0.908,
+      "step": 12000
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 1.0925297113752123e-06,
+      "loss": 0.8921,
+      "step": 12500
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.3023733455043359,
+      "eval_loss": 2.518986701965332,
+      "eval_runtime": 11.4293,
+      "eval_samples_per_second": 383.399,
+      "eval_steps_per_second": 47.947,
+      "step": 12780
+    }
+  ],
+  "max_steps": 12780,
+  "num_train_epochs": 10,
+  "total_flos": 418971279265446.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a21031e6690f1b602be5eaabc87b8be44d1fa879d11f0d31295baaffed38a8d3
+size 3311