Training in progress, step 200

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 1.6,
+    "eval_loss": 0.29148319363594055,
+    "eval_runtime": 3.519,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 568.336,
+    "eval_steps_per_second": 35.521,
+    "total_flos": 726030019897056.0,
+    "train_loss": 0.32366644477844236,
+    "train_runtime": 436.683,
+    "train_samples": 20000,
+    "train_samples_per_second": 73.28,
+    "train_steps_per_second": 4.58
+}

config.json

@@ -1,10 +1,12 @@
 {
-  "_name_or_path": "microsoft/Multilingual-MiniLM-L12-H384",
+  "_name_or_path": "nreimers/mMiniLMv2-L6-H384-distilled-from-XLMR-Large",
   "architectures": [
-    "BertForSequenceClassification"
+    "XLMRobertaForSequenceClassification"
   ],
   "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": 384,
@@ -16,18 +18,17 @@
   "label2id": {
     "LABEL_0": 0
   },
-  "layer_norm_eps": 1e-12,
-  "max_position_embeddings": 512,
-  "model_type": "bert",
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "xlm-roberta",
   "num_attention_heads": 12,
-  "num_hidden_layers": 12,
-  "pad_token_id": 0,
+  "num_hidden_layers": 6,
+  "pad_token_id": 1,
   "position_embedding_type": "absolute",
   "problem_type": "multi_label_classification",
-  "tokenizer_class": "XLMRobertaTokenizer",
   "torch_dtype": "float32",
   "transformers_version": "4.35.2",
-  "type_vocab_size": 2,
+  "type_vocab_size": 1,
   "use_cache": true,
-  "vocab_size": 250037
+  "vocab_size": 250002
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.6,
+    "eval_loss": 0.29148319363594055,
+    "eval_runtime": 3.519,
+    "eval_samples": 2000,
+    "eval_samples_per_second": 568.336,
+    "eval_steps_per_second": 35.521
+}

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:123f34587638ab07c3c6a77ca807b5c73a25a2702528d7ac871c240357815314
-size 470640124
+oid sha256:90278f7b011debb895c4fbc46931f27f6c616ef4d8ca00fac6f7c9481a0b5e0f
+size 427989732

tokenizer.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f406e7c7e1298a6ea5497766446592bc718ea5f830fc0da5e6c7d448909a17c0
-size 17098270
+oid sha256:f59925fcb90c92b894cb93e51bb9b4a6105c5c249fe54ce1c704420ac39b81af
+size 17082756

tokenizer_config.json

@@ -46,10 +46,9 @@
   "cls_token": "<s>",
   "eos_token": "</s>",
   "mask_token": "<mask>",
-  "model_max_length": 1000000000000000019884624838656,
+  "model_max_length": 512,
   "pad_token": "<pad>",
   "sep_token": "</s>",
-  "sp_model_kwargs": {},
   "tokenizer_class": "XLMRobertaTokenizer",
   "unk_token": "<unk>"
 }

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.6,
+    "total_flos": 726030019897056.0,
+    "train_loss": 0.32366644477844236,
+    "train_runtime": 436.683,
+    "train_samples": 20000,
+    "train_samples_per_second": 73.28,
+    "train_steps_per_second": 4.58
+}

trainer_state.json

@@ -0,0 +1,186 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6,
+  "eval_steps": 200,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.175000000000001e-06,
+      "loss": 0.4046,
+      "step": 150
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.36089497804641724,
+      "eval_runtime": 3.5302,
+      "eval_samples_per_second": 566.546,
+      "eval_steps_per_second": 35.409,
+      "step": 200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.925000000000001e-06,
+      "loss": 0.3704,
+      "step": 300
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.3112202286720276,
+      "eval_runtime": 3.5156,
+      "eval_samples_per_second": 568.898,
+      "eval_steps_per_second": 35.556,
+      "step": 400
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 6.683333333333334e-06,
+      "loss": 0.3767,
+      "step": 450
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.4333333333333335e-06,
+      "loss": 0.3404,
+      "step": 600
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 0.31385645270347595,
+      "eval_runtime": 3.5516,
+      "eval_samples_per_second": 563.134,
+      "eval_steps_per_second": 35.196,
+      "step": 600
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.183333333333334e-06,
+      "loss": 0.3208,
+      "step": 750
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.29552698135375977,
+      "eval_runtime": 3.5281,
+      "eval_samples_per_second": 566.883,
+      "eval_steps_per_second": 35.43,
+      "step": 800
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 0.3139,
+      "step": 900
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.2942730486392975,
+      "eval_runtime": 3.5869,
+      "eval_samples_per_second": 557.582,
+      "eval_steps_per_second": 34.849,
+      "step": 1000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6833333333333335e-06,
+      "loss": 0.3271,
+      "step": 1050
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.333333333333334e-07,
+      "loss": 0.3061,
+      "step": 1200
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.2933160662651062,
+      "eval_runtime": 3.5585,
+      "eval_samples_per_second": 562.034,
+      "eval_steps_per_second": 35.127,
+      "step": 1200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0,
+      "loss": 0.2857,
+      "step": 1350
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 0.29148319363594055,
+      "eval_runtime": 3.5159,
+      "eval_samples_per_second": 568.844,
+      "eval_steps_per_second": 35.553,
+      "step": 1400
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0,
+      "loss": 0.3086,
+      "step": 1500
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 0.29148319363594055,
+      "eval_runtime": 3.6912,
+      "eval_samples_per_second": 541.829,
+      "eval_steps_per_second": 33.864,
+      "step": 1600
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0,
+      "loss": 0.2944,
+      "step": 1650
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0,
+      "loss": 0.283,
+      "step": 1800
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 0.29148319363594055,
+      "eval_runtime": 3.5074,
+      "eval_samples_per_second": 570.226,
+      "eval_steps_per_second": 35.639,
+      "step": 1800
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0,
+      "loss": 0.2855,
+      "step": 1950
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 0.29148319363594055,
+      "eval_runtime": 3.719,
+      "eval_samples_per_second": 537.774,
+      "eval_steps_per_second": 33.611,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6,
+      "step": 2000,
+      "total_flos": 726030019897056.0,
+      "train_loss": 0.32366644477844236,
+      "train_runtime": 436.683,
+      "train_samples_per_second": 73.28,
+      "train_steps_per_second": 4.58
+    }
+  ],
+  "logging_steps": 150,
+  "max_steps": 2000,
+  "num_train_epochs": 2,
+  "save_steps": 200,
+  "total_flos": 726030019897056.0,
+  "trial_name": null,
+  "trial_params": null
+}