Training in progress, step 1924

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

adapter_config.json

@@ -0,0 +1,18 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "feedforward_modules": "down_proj",
+  "inference_mode": true,
+  "init_ia3_weights": true,
+  "modules_to_save": null,
+  "peft_type": "IA3",
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "k_proj",
+    "down_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9410ddf90dfcaa463f8c701345f9007d2f2fe4351ab0fe99a0af059b76206908
+size 798032

special_tokens_map.json

@@ -0,0 +1,26 @@
+{
+  "additional_special_tokens": [
+    {
+      "content": "<|eom_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false
+    }
+  ],
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|eot_id|>"
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c1dcab308e7cf5970ea38815e0a62887d705c5b436f869ca27a5dcdd40c36a6
+size 17210148

tokenizer_config.json

@@ -0,0 +1,2077 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128256": {
+      "content": "<|eom_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|eom_id|>"
+  ],
+  "bos_token": "<|begin_of_text|>",
+  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|eot_id|>",
+  "extra_special_tokens": {},
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|eot_id|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "PreTrainedTokenizer"
+}

train.yaml

@@ -0,0 +1,61 @@
+seed: 456
+
+### model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
+flash_attn: auto
+use_cache: false
+
+### method
+stage: sft
+do_train: true
+finetuning_type: ia3
+task_type: CAUSAL_LM
+target_modules:
+- k_proj
+- v_proj
+- down_proj
+feedforward_modules: down_proj
+
+### dataset
+dataset: cola
+template: llama3
+cutoff_len: 2048
+overwrite_cache: true
+preprocessing_num_workers: 8
+dataloader_num_workers: 2
+packing: false
+
+### output
+output_dir: saves_multiple/ia3/llama-3-8b-instruct/train_cola_456_1760637821
+logging_steps: 5
+save_steps: 0.05
+overwrite_output_dir: true
+save_only_model: false
+plot_loss: true
+include_num_input_tokens_seen: true
+push_to_hub: true
+push_to_hub_organization: rbelanec
+load_best_model_at_end: true
+save_total_limit: 1
+
+### train
+per_device_train_batch_size: 4
+learning_rate: 5.0e-5
+num_train_epochs: 20
+weight_decay: 1.0e-5
+lr_scheduler_type: cosine
+bf16: true
+ddp_timeout: 180000000
+resume_from_checkpoint: null
+warmup_ratio: 0.1
+optim: adamw_torch
+report_to:
+- wandb
+run_name: ia3_llama-3-8b-instruct_train_cola_456_1760637821
+
+### eval
+per_device_eval_batch_size: 4
+eval_strategy: steps
+eval_steps: 0.05
+val_size: 0.1

trainer_log.jsonl

@@ -0,0 +1,390 @@
+{"current_steps": 5, "total_steps": 38480, "loss": 2.1137, "lr": 5.197505197505198e-08, "epoch": 0.002598752598752599, "percentage": 0.01, "elapsed_time": "0:00:01", "remaining_time": "2:28:14", "throughput": 830.48, "total_tokens": 960}
+{"current_steps": 10, "total_steps": 38480, "loss": 1.0815, "lr": 1.1694386694386695e-07, "epoch": 0.005197505197505198, "percentage": 0.03, "elapsed_time": "0:00:01", "remaining_time": "1:59:18", "throughput": 1049.05, "total_tokens": 1952}
+{"current_steps": 15, "total_steps": 38480, "loss": 1.681, "lr": 1.8191268191268192e-07, "epoch": 0.007796257796257797, "percentage": 0.04, "elapsed_time": "0:00:02", "remaining_time": "1:49:38", "throughput": 1160.02, "total_tokens": 2976}
+{"current_steps": 20, "total_steps": 38480, "loss": 1.9192, "lr": 2.468814968814969e-07, "epoch": 0.010395010395010396, "percentage": 0.05, "elapsed_time": "0:00:03", "remaining_time": "1:44:47", "throughput": 1213.53, "total_tokens": 3968}
+{"current_steps": 25, "total_steps": 38480, "loss": 1.7691, "lr": 3.1185031185031186e-07, "epoch": 0.012993762993762994, "percentage": 0.06, "elapsed_time": "0:00:03", "remaining_time": "1:41:53", "throughput": 1223.79, "total_tokens": 4864}
+{"current_steps": 30, "total_steps": 38480, "loss": 1.1956, "lr": 3.768191268191269e-07, "epoch": 0.015592515592515593, "percentage": 0.08, "elapsed_time": "0:00:04", "remaining_time": "1:39:53", "throughput": 1238.63, "total_tokens": 5792}
+{"current_steps": 35, "total_steps": 38480, "loss": 1.5342, "lr": 4.4178794178794185e-07, "epoch": 0.018191268191268192, "percentage": 0.09, "elapsed_time": "0:00:05", "remaining_time": "1:38:30", "throughput": 1260.81, "total_tokens": 6784}
+{"current_steps": 40, "total_steps": 38480, "loss": 1.3079, "lr": 5.067567567567568e-07, "epoch": 0.02079002079002079, "percentage": 0.1, "elapsed_time": "0:00:06", "remaining_time": "1:37:19", "throughput": 1279.59, "total_tokens": 7776}
+{"current_steps": 45, "total_steps": 38480, "loss": 1.1775, "lr": 5.717255717255718e-07, "epoch": 0.02338877338877339, "percentage": 0.12, "elapsed_time": "0:00:06", "remaining_time": "1:36:23", "throughput": 1290.1, "total_tokens": 8736}
+{"current_steps": 50, "total_steps": 38480, "loss": 1.6009, "lr": 6.366943866943868e-07, "epoch": 0.02598752598752599, "percentage": 0.13, "elapsed_time": "0:00:07", "remaining_time": "1:35:38", "throughput": 1294.28, "total_tokens": 9664}
+{"current_steps": 55, "total_steps": 38480, "loss": 1.3078, "lr": 7.016632016632017e-07, "epoch": 0.028586278586278588, "percentage": 0.14, "elapsed_time": "0:00:08", "remaining_time": "1:35:04", "throughput": 1305.12, "total_tokens": 10656}
+{"current_steps": 60, "total_steps": 38480, "loss": 1.3145, "lr": 7.666320166320167e-07, "epoch": 0.031185031185031187, "percentage": 0.16, "elapsed_time": "0:00:08", "remaining_time": "1:34:35", "throughput": 1306.85, "total_tokens": 11584}
+{"current_steps": 65, "total_steps": 38480, "loss": 1.5184, "lr": 8.316008316008317e-07, "epoch": 0.033783783783783786, "percentage": 0.17, "elapsed_time": "0:00:09", "remaining_time": "1:34:11", "throughput": 1305.04, "total_tokens": 12480}
+{"current_steps": 70, "total_steps": 38480, "loss": 1.0893, "lr": 8.965696465696466e-07, "epoch": 0.036382536382536385, "percentage": 0.18, "elapsed_time": "0:00:10", "remaining_time": "1:33:51", "throughput": 1303.28, "total_tokens": 13376}
+{"current_steps": 75, "total_steps": 38480, "loss": 1.3677, "lr": 9.615384615384617e-07, "epoch": 0.03898128898128898, "percentage": 0.19, "elapsed_time": "0:00:10", "remaining_time": "1:33:34", "throughput": 1301.6, "total_tokens": 14272}
+{"current_steps": 80, "total_steps": 38480, "loss": 1.3394, "lr": 1.0265072765072766e-06, "epoch": 0.04158004158004158, "percentage": 0.21, "elapsed_time": "0:00:11", "remaining_time": "1:33:19", "throughput": 1303.03, "total_tokens": 15200}
+{"current_steps": 85, "total_steps": 38480, "loss": 1.7587, "lr": 1.0914760914760916e-06, "epoch": 0.04417879417879418, "percentage": 0.22, "elapsed_time": "0:00:12", "remaining_time": "1:33:05", "throughput": 1304.39, "total_tokens": 16128}
+{"current_steps": 90, "total_steps": 38480, "loss": 1.3611, "lr": 1.1564449064449065e-06, "epoch": 0.04677754677754678, "percentage": 0.23, "elapsed_time": "0:00:13", "remaining_time": "1:32:52", "throughput": 1308.13, "total_tokens": 17088}
+{"current_steps": 95, "total_steps": 38480, "loss": 1.5119, "lr": 1.2214137214137216e-06, "epoch": 0.04937629937629938, "percentage": 0.25, "elapsed_time": "0:00:13", "remaining_time": "1:32:40", "throughput": 1309.23, "total_tokens": 18016}
+{"current_steps": 100, "total_steps": 38480, "loss": 1.3283, "lr": 1.2863825363825364e-06, "epoch": 0.05197505197505198, "percentage": 0.26, "elapsed_time": "0:00:14", "remaining_time": "1:32:28", "throughput": 1312.5, "total_tokens": 18976}
+{"current_steps": 105, "total_steps": 38480, "loss": 1.2739, "lr": 1.3513513513513515e-06, "epoch": 0.05457380457380458, "percentage": 0.27, "elapsed_time": "0:00:15", "remaining_time": "1:32:21", "throughput": 1314.76, "total_tokens": 19936}
+{"current_steps": 110, "total_steps": 38480, "loss": 1.4597, "lr": 1.4163201663201664e-06, "epoch": 0.057172557172557176, "percentage": 0.29, "elapsed_time": "0:00:15", "remaining_time": "1:32:16", "throughput": 1320.5, "total_tokens": 20960}
+{"current_steps": 115, "total_steps": 38480, "loss": 1.6943, "lr": 1.4812889812889814e-06, "epoch": 0.059771309771309775, "percentage": 0.3, "elapsed_time": "0:00:16", "remaining_time": "1:32:08", "throughput": 1324.57, "total_tokens": 21952}
+{"current_steps": 120, "total_steps": 38480, "loss": 1.5935, "lr": 1.5462577962577965e-06, "epoch": 0.062370062370062374, "percentage": 0.31, "elapsed_time": "0:00:17", "remaining_time": "1:32:01", "throughput": 1330.09, "total_tokens": 22976}
+{"current_steps": 125, "total_steps": 38480, "loss": 1.4063, "lr": 1.6112266112266114e-06, "epoch": 0.06496881496881497, "percentage": 0.32, "elapsed_time": "0:00:17", "remaining_time": "1:31:55", "throughput": 1331.69, "total_tokens": 23936}
+{"current_steps": 130, "total_steps": 38480, "loss": 1.477, "lr": 1.6761954261954265e-06, "epoch": 0.06756756756756757, "percentage": 0.34, "elapsed_time": "0:00:18", "remaining_time": "1:31:49", "throughput": 1333.15, "total_tokens": 24896}
+{"current_steps": 135, "total_steps": 38480, "loss": 1.3631, "lr": 1.7411642411642413e-06, "epoch": 0.07016632016632017, "percentage": 0.35, "elapsed_time": "0:00:19", "remaining_time": "1:31:43", "throughput": 1332.81, "total_tokens": 25824}
+{"current_steps": 140, "total_steps": 38480, "loss": 1.3002, "lr": 1.8061330561330564e-06, "epoch": 0.07276507276507277, "percentage": 0.36, "elapsed_time": "0:00:20", "remaining_time": "1:31:38", "throughput": 1330.91, "total_tokens": 26720}
+{"current_steps": 145, "total_steps": 38480, "loss": 1.3947, "lr": 1.8711018711018713e-06, "epoch": 0.07536382536382537, "percentage": 0.38, "elapsed_time": "0:00:20", "remaining_time": "1:31:31", "throughput": 1327.94, "total_tokens": 27584}
+{"current_steps": 150, "total_steps": 38480, "loss": 1.4056, "lr": 1.9360706860706863e-06, "epoch": 0.07796257796257797, "percentage": 0.39, "elapsed_time": "0:00:21", "remaining_time": "1:31:27", "throughput": 1330.7, "total_tokens": 28576}
+{"current_steps": 155, "total_steps": 38480, "loss": 1.3329, "lr": 2.0010395010395014e-06, "epoch": 0.08056133056133057, "percentage": 0.4, "elapsed_time": "0:00:22", "remaining_time": "1:31:22", "throughput": 1332.05, "total_tokens": 29536}
+{"current_steps": 160, "total_steps": 38480, "loss": 1.1216, "lr": 2.066008316008316e-06, "epoch": 0.08316008316008316, "percentage": 0.42, "elapsed_time": "0:00:22", "remaining_time": "1:31:17", "throughput": 1330.59, "total_tokens": 30432}
+{"current_steps": 165, "total_steps": 38480, "loss": 1.189, "lr": 2.130977130977131e-06, "epoch": 0.08575883575883576, "percentage": 0.43, "elapsed_time": "0:00:23", "remaining_time": "1:31:13", "throughput": 1330.46, "total_tokens": 31360}
+{"current_steps": 170, "total_steps": 38480, "loss": 1.3894, "lr": 2.195945945945946e-06, "epoch": 0.08835758835758836, "percentage": 0.44, "elapsed_time": "0:00:24", "remaining_time": "1:31:08", "throughput": 1329.23, "total_tokens": 32256}
+{"current_steps": 175, "total_steps": 38480, "loss": 1.5356, "lr": 2.2609147609147613e-06, "epoch": 0.09095634095634096, "percentage": 0.45, "elapsed_time": "0:00:24", "remaining_time": "1:31:03", "throughput": 1326.81, "total_tokens": 33120}
+{"current_steps": 180, "total_steps": 38480, "loss": 1.7132, "lr": 2.325883575883576e-06, "epoch": 0.09355509355509356, "percentage": 0.47, "elapsed_time": "0:00:25", "remaining_time": "1:31:00", "throughput": 1326.78, "total_tokens": 34048}
+{"current_steps": 185, "total_steps": 38480, "loss": 1.3681, "lr": 2.390852390852391e-06, "epoch": 0.09615384615384616, "percentage": 0.48, "elapsed_time": "0:00:26", "remaining_time": "1:30:56", "throughput": 1326.86, "total_tokens": 34976}
+{"current_steps": 190, "total_steps": 38480, "loss": 1.3093, "lr": 2.455821205821206e-06, "epoch": 0.09875259875259876, "percentage": 0.49, "elapsed_time": "0:00:27", "remaining_time": "1:30:52", "throughput": 1329.3, "total_tokens": 35968}
+{"current_steps": 195, "total_steps": 38480, "loss": 1.1733, "lr": 2.520790020790021e-06, "epoch": 0.10135135135135136, "percentage": 0.51, "elapsed_time": "0:00:27", "remaining_time": "1:30:49", "throughput": 1331.61, "total_tokens": 36960}
+{"current_steps": 200, "total_steps": 38480, "loss": 1.5229, "lr": 2.5857588357588362e-06, "epoch": 0.10395010395010396, "percentage": 0.52, "elapsed_time": "0:00:28", "remaining_time": "1:30:45", "throughput": 1331.73, "total_tokens": 37888}
+{"current_steps": 205, "total_steps": 38480, "loss": 1.1705, "lr": 2.650727650727651e-06, "epoch": 0.10654885654885655, "percentage": 0.53, "elapsed_time": "0:00:29", "remaining_time": "1:30:41", "throughput": 1334.11, "total_tokens": 38880}
+{"current_steps": 210, "total_steps": 38480, "loss": 1.7054, "lr": 2.715696465696466e-06, "epoch": 0.10914760914760915, "percentage": 0.55, "elapsed_time": "0:00:29", "remaining_time": "1:30:37", "throughput": 1335.25, "total_tokens": 39840}
+{"current_steps": 215, "total_steps": 38480, "loss": 1.3018, "lr": 2.780665280665281e-06, "epoch": 0.11174636174636175, "percentage": 0.56, "elapsed_time": "0:00:30", "remaining_time": "1:30:35", "throughput": 1333.94, "total_tokens": 40736}
+{"current_steps": 220, "total_steps": 38480, "loss": 1.2685, "lr": 2.845634095634096e-06, "epoch": 0.11434511434511435, "percentage": 0.57, "elapsed_time": "0:00:31", "remaining_time": "1:30:31", "throughput": 1335.99, "total_tokens": 41728}
+{"current_steps": 225, "total_steps": 38480, "loss": 2.0038, "lr": 2.9106029106029107e-06, "epoch": 0.11694386694386695, "percentage": 0.58, "elapsed_time": "0:00:31", "remaining_time": "1:30:28", "throughput": 1335.95, "total_tokens": 42656}
+{"current_steps": 230, "total_steps": 38480, "loss": 1.4704, "lr": 2.975571725571726e-06, "epoch": 0.11954261954261955, "percentage": 0.6, "elapsed_time": "0:00:32", "remaining_time": "1:30:25", "throughput": 1336.91, "total_tokens": 43616}
+{"current_steps": 235, "total_steps": 38480, "loss": 1.1395, "lr": 3.040540540540541e-06, "epoch": 0.12214137214137215, "percentage": 0.61, "elapsed_time": "0:00:33", "remaining_time": "1:30:23", "throughput": 1336.74, "total_tokens": 44544}
+{"current_steps": 240, "total_steps": 38480, "loss": 1.3822, "lr": 3.105509355509356e-06, "epoch": 0.12474012474012475, "percentage": 0.62, "elapsed_time": "0:00:34", "remaining_time": "1:30:20", "throughput": 1335.73, "total_tokens": 45440}
+{"current_steps": 245, "total_steps": 38480, "loss": 1.3046, "lr": 3.1704781704781706e-06, "epoch": 0.12733887733887733, "percentage": 0.64, "elapsed_time": "0:00:34", "remaining_time": "1:30:17", "throughput": 1336.57, "total_tokens": 46400}
+{"current_steps": 250, "total_steps": 38480, "loss": 1.3256, "lr": 3.2354469854469853e-06, "epoch": 0.12993762993762994, "percentage": 0.65, "elapsed_time": "0:00:35", "remaining_time": "1:30:15", "throughput": 1337.32, "total_tokens": 47360}
+{"current_steps": 255, "total_steps": 38480, "loss": 1.2828, "lr": 3.3004158004158008e-06, "epoch": 0.13253638253638253, "percentage": 0.66, "elapsed_time": "0:00:36", "remaining_time": "1:30:12", "throughput": 1339.05, "total_tokens": 48352}
+{"current_steps": 260, "total_steps": 38480, "loss": 1.5133, "lr": 3.3653846153846154e-06, "epoch": 0.13513513513513514, "percentage": 0.68, "elapsed_time": "0:00:36", "remaining_time": "1:30:10", "throughput": 1338.86, "total_tokens": 49280}
+{"current_steps": 265, "total_steps": 38480, "loss": 1.3136, "lr": 3.430353430353431e-06, "epoch": 0.13773388773388773, "percentage": 0.69, "elapsed_time": "0:00:37", "remaining_time": "1:30:08", "throughput": 1338.73, "total_tokens": 50208}
+{"current_steps": 270, "total_steps": 38480, "loss": 1.6679, "lr": 3.495322245322245e-06, "epoch": 0.14033264033264034, "percentage": 0.7, "elapsed_time": "0:00:38", "remaining_time": "1:30:06", "throughput": 1338.59, "total_tokens": 51136}
+{"current_steps": 275, "total_steps": 38480, "loss": 1.3045, "lr": 3.5602910602910606e-06, "epoch": 0.14293139293139293, "percentage": 0.71, "elapsed_time": "0:00:38", "remaining_time": "1:30:04", "throughput": 1338.29, "total_tokens": 52064}
+{"current_steps": 280, "total_steps": 38480, "loss": 1.4569, "lr": 3.6252598752598753e-06, "epoch": 0.14553014553014554, "percentage": 0.73, "elapsed_time": "0:00:39", "remaining_time": "1:30:02", "throughput": 1338.9, "total_tokens": 53024}
+{"current_steps": 285, "total_steps": 38480, "loss": 1.1014, "lr": 3.6902286902286908e-06, "epoch": 0.14812889812889812, "percentage": 0.74, "elapsed_time": "0:00:40", "remaining_time": "1:30:01", "throughput": 1339.48, "total_tokens": 53984}
+{"current_steps": 290, "total_steps": 38480, "loss": 1.3949, "lr": 3.7551975051975054e-06, "epoch": 0.15072765072765074, "percentage": 0.75, "elapsed_time": "0:00:41", "remaining_time": "1:29:59", "throughput": 1340.86, "total_tokens": 54976}
+{"current_steps": 295, "total_steps": 38480, "loss": 1.2977, "lr": 3.820166320166321e-06, "epoch": 0.15332640332640332, "percentage": 0.77, "elapsed_time": "0:00:41", "remaining_time": "1:29:57", "throughput": 1340.59, "total_tokens": 55904}
+{"current_steps": 300, "total_steps": 38480, "loss": 1.7146, "lr": 3.885135135135135e-06, "epoch": 0.15592515592515593, "percentage": 0.78, "elapsed_time": "0:00:42", "remaining_time": "1:29:55", "throughput": 1341.18, "total_tokens": 56864}
+{"current_steps": 305, "total_steps": 38480, "loss": 1.4164, "lr": 3.95010395010395e-06, "epoch": 0.15852390852390852, "percentage": 0.79, "elapsed_time": "0:00:43", "remaining_time": "1:29:54", "throughput": 1340.15, "total_tokens": 57760}
+{"current_steps": 310, "total_steps": 38480, "loss": 1.3343, "lr": 4.015072765072765e-06, "epoch": 0.16112266112266113, "percentage": 0.81, "elapsed_time": "0:00:43", "remaining_time": "1:29:52", "throughput": 1340.77, "total_tokens": 58720}
+{"current_steps": 315, "total_steps": 38480, "loss": 1.6583, "lr": 4.08004158004158e-06, "epoch": 0.16372141372141372, "percentage": 0.82, "elapsed_time": "0:00:44", "remaining_time": "1:29:50", "throughput": 1341.29, "total_tokens": 59680}
+{"current_steps": 320, "total_steps": 38480, "loss": 1.1167, "lr": 4.1450103950103955e-06, "epoch": 0.16632016632016633, "percentage": 0.83, "elapsed_time": "0:00:45", "remaining_time": "1:29:49", "throughput": 1342.39, "total_tokens": 60672}
+{"current_steps": 325, "total_steps": 38480, "loss": 1.32, "lr": 4.2099792099792105e-06, "epoch": 0.16891891891891891, "percentage": 0.84, "elapsed_time": "0:00:45", "remaining_time": "1:29:48", "throughput": 1342.19, "total_tokens": 61600}
+{"current_steps": 330, "total_steps": 38480, "loss": 1.4241, "lr": 4.274948024948025e-06, "epoch": 0.17151767151767153, "percentage": 0.86, "elapsed_time": "0:00:46", "remaining_time": "1:29:46", "throughput": 1342.02, "total_tokens": 62528}
+{"current_steps": 335, "total_steps": 38480, "loss": 1.2792, "lr": 4.339916839916841e-06, "epoch": 0.1741164241164241, "percentage": 0.87, "elapsed_time": "0:00:47", "remaining_time": "1:29:44", "throughput": 1341.18, "total_tokens": 63424}
+{"current_steps": 340, "total_steps": 38480, "loss": 1.5734, "lr": 4.404885654885655e-06, "epoch": 0.17671517671517672, "percentage": 0.88, "elapsed_time": "0:00:47", "remaining_time": "1:29:42", "throughput": 1341.71, "total_tokens": 64384}
+{"current_steps": 345, "total_steps": 38480, "loss": 1.3601, "lr": 4.46985446985447e-06, "epoch": 0.1793139293139293, "percentage": 0.9, "elapsed_time": "0:00:48", "remaining_time": "1:29:41", "throughput": 1342.88, "total_tokens": 65376}
+{"current_steps": 350, "total_steps": 38480, "loss": 1.2367, "lr": 4.534823284823285e-06, "epoch": 0.18191268191268192, "percentage": 0.91, "elapsed_time": "0:00:49", "remaining_time": "1:29:39", "throughput": 1342.74, "total_tokens": 66304}
+{"current_steps": 355, "total_steps": 38480, "loss": 1.4771, "lr": 4.5997920997921e-06, "epoch": 0.1845114345114345, "percentage": 0.92, "elapsed_time": "0:00:50", "remaining_time": "1:29:38", "throughput": 1342.55, "total_tokens": 67232}
+{"current_steps": 360, "total_steps": 38480, "loss": 1.02, "lr": 4.664760914760915e-06, "epoch": 0.18711018711018712, "percentage": 0.94, "elapsed_time": "0:00:50", "remaining_time": "1:29:36", "throughput": 1342.4, "total_tokens": 68160}
+{"current_steps": 365, "total_steps": 38480, "loss": 1.6013, "lr": 4.72972972972973e-06, "epoch": 0.1897089397089397, "percentage": 0.95, "elapsed_time": "0:00:51", "remaining_time": "1:29:34", "throughput": 1341.69, "total_tokens": 69056}
+{"current_steps": 370, "total_steps": 38480, "loss": 1.8517, "lr": 4.7946985446985445e-06, "epoch": 0.19230769230769232, "percentage": 0.96, "elapsed_time": "0:00:52", "remaining_time": "1:29:32", "throughput": 1343.51, "total_tokens": 70080}
+{"current_steps": 375, "total_steps": 38480, "loss": 1.1191, "lr": 4.8596673596673604e-06, "epoch": 0.1949064449064449, "percentage": 0.97, "elapsed_time": "0:00:52", "remaining_time": "1:29:30", "throughput": 1344.09, "total_tokens": 71040}
+{"current_steps": 380, "total_steps": 38480, "loss": 1.3228, "lr": 4.924636174636175e-06, "epoch": 0.19750519750519752, "percentage": 0.99, "elapsed_time": "0:00:53", "remaining_time": "1:29:28", "throughput": 1344.05, "total_tokens": 71968}
+{"current_steps": 385, "total_steps": 38480, "loss": 1.6021, "lr": 4.98960498960499e-06, "epoch": 0.2001039501039501, "percentage": 1.0, "elapsed_time": "0:00:54", "remaining_time": "1:29:26", "throughput": 1344.01, "total_tokens": 72896}
+{"current_steps": 390, "total_steps": 38480, "loss": 1.13, "lr": 5.054573804573805e-06, "epoch": 0.20270270270270271, "percentage": 1.01, "elapsed_time": "0:00:54", "remaining_time": "1:29:24", "throughput": 1344.51, "total_tokens": 73856}
+{"current_steps": 395, "total_steps": 38480, "loss": 1.0121, "lr": 5.11954261954262e-06, "epoch": 0.2053014553014553, "percentage": 1.03, "elapsed_time": "0:00:55", "remaining_time": "1:29:23", "throughput": 1345.01, "total_tokens": 74816}
+{"current_steps": 400, "total_steps": 38480, "loss": 1.3348, "lr": 5.184511434511435e-06, "epoch": 0.2079002079002079, "percentage": 1.04, "elapsed_time": "0:00:56", "remaining_time": "1:29:21", "throughput": 1345.52, "total_tokens": 75776}
+{"current_steps": 405, "total_steps": 38480, "loss": 1.3664, "lr": 5.24948024948025e-06, "epoch": 0.2104989604989605, "percentage": 1.05, "elapsed_time": "0:00:57", "remaining_time": "1:29:19", "throughput": 1345.46, "total_tokens": 76704}
+{"current_steps": 410, "total_steps": 38480, "loss": 1.3917, "lr": 5.314449064449064e-06, "epoch": 0.2130977130977131, "percentage": 1.07, "elapsed_time": "0:00:57", "remaining_time": "1:29:17", "throughput": 1344.29, "total_tokens": 77568}
+{"current_steps": 415, "total_steps": 38480, "loss": 1.2604, "lr": 5.37941787941788e-06, "epoch": 0.2156964656964657, "percentage": 1.08, "elapsed_time": "0:00:58", "remaining_time": "1:29:16", "throughput": 1344.25, "total_tokens": 78496}
+{"current_steps": 420, "total_steps": 38480, "loss": 1.364, "lr": 5.444386694386694e-06, "epoch": 0.2182952182952183, "percentage": 1.09, "elapsed_time": "0:00:59", "remaining_time": "1:29:14", "throughput": 1344.64, "total_tokens": 79456}
+{"current_steps": 425, "total_steps": 38480, "loss": 1.1632, "lr": 5.50935550935551e-06, "epoch": 0.2208939708939709, "percentage": 1.1, "elapsed_time": "0:00:59", "remaining_time": "1:29:13", "throughput": 1344.59, "total_tokens": 80384}
+{"current_steps": 430, "total_steps": 38480, "loss": 1.7891, "lr": 5.5743243243243245e-06, "epoch": 0.2234927234927235, "percentage": 1.12, "elapsed_time": "0:01:00", "remaining_time": "1:29:12", "throughput": 1343.81, "total_tokens": 81280}
+{"current_steps": 435, "total_steps": 38480, "loss": 1.5949, "lr": 5.63929313929314e-06, "epoch": 0.2260914760914761, "percentage": 1.13, "elapsed_time": "0:01:01", "remaining_time": "1:29:10", "throughput": 1344.77, "total_tokens": 82272}
+{"current_steps": 440, "total_steps": 38480, "loss": 1.4827, "lr": 5.704261954261955e-06, "epoch": 0.2286902286902287, "percentage": 1.14, "elapsed_time": "0:01:01", "remaining_time": "1:29:09", "throughput": 1345.74, "total_tokens": 83264}
+{"current_steps": 445, "total_steps": 38480, "loss": 1.5216, "lr": 5.76923076923077e-06, "epoch": 0.2312889812889813, "percentage": 1.16, "elapsed_time": "0:01:02", "remaining_time": "1:29:08", "throughput": 1346.01, "total_tokens": 84224}
+{"current_steps": 450, "total_steps": 38480, "loss": 1.6149, "lr": 5.834199584199585e-06, "epoch": 0.2338877338877339, "percentage": 1.17, "elapsed_time": "0:01:03", "remaining_time": "1:29:07", "throughput": 1345.82, "total_tokens": 85152}
+{"current_steps": 455, "total_steps": 38480, "loss": 1.5695, "lr": 5.8991683991684e-06, "epoch": 0.23648648648648649, "percentage": 1.18, "elapsed_time": "0:01:03", "remaining_time": "1:29:05", "throughput": 1345.65, "total_tokens": 86080}
+{"current_steps": 460, "total_steps": 38480, "loss": 1.1811, "lr": 5.964137214137214e-06, "epoch": 0.2390852390852391, "percentage": 1.2, "elapsed_time": "0:01:04", "remaining_time": "1:29:04", "throughput": 1345.96, "total_tokens": 87040}
+{"current_steps": 465, "total_steps": 38480, "loss": 1.3844, "lr": 6.02910602910603e-06, "epoch": 0.24168399168399168, "percentage": 1.21, "elapsed_time": "0:01:05", "remaining_time": "1:29:03", "throughput": 1345.79, "total_tokens": 87968}
+{"current_steps": 470, "total_steps": 38480, "loss": 1.2354, "lr": 6.094074844074844e-06, "epoch": 0.2442827442827443, "percentage": 1.22, "elapsed_time": "0:01:06", "remaining_time": "1:29:03", "throughput": 1346.97, "total_tokens": 88992}
+{"current_steps": 475, "total_steps": 38480, "loss": 1.7011, "lr": 6.159043659043659e-06, "epoch": 0.24688149688149688, "percentage": 1.23, "elapsed_time": "0:01:06", "remaining_time": "1:29:02", "throughput": 1347.73, "total_tokens": 89984}
+{"current_steps": 480, "total_steps": 38480, "loss": 1.3419, "lr": 6.2240124740124744e-06, "epoch": 0.2494802494802495, "percentage": 1.25, "elapsed_time": "0:01:07", "remaining_time": "1:29:00", "throughput": 1349.04, "total_tokens": 91008}
+{"current_steps": 485, "total_steps": 38480, "loss": 1.665, "lr": 6.2889812889812895e-06, "epoch": 0.2520790020790021, "percentage": 1.26, "elapsed_time": "0:01:08", "remaining_time": "1:28:59", "throughput": 1349.77, "total_tokens": 92000}
+{"current_steps": 490, "total_steps": 38480, "loss": 1.3007, "lr": 6.353950103950105e-06, "epoch": 0.25467775467775466, "percentage": 1.27, "elapsed_time": "0:01:08", "remaining_time": "1:28:58", "throughput": 1349.57, "total_tokens": 92928}
+{"current_steps": 495, "total_steps": 38480, "loss": 1.9481, "lr": 6.418918918918919e-06, "epoch": 0.25727650727650725, "percentage": 1.29, "elapsed_time": "0:01:09", "remaining_time": "1:28:57", "throughput": 1350.73, "total_tokens": 93952}
+{"current_steps": 500, "total_steps": 38480, "loss": 1.4173, "lr": 6.483887733887734e-06, "epoch": 0.2598752598752599, "percentage": 1.3, "elapsed_time": "0:01:10", "remaining_time": "1:28:56", "throughput": 1350.43, "total_tokens": 94880}
+{"current_steps": 505, "total_steps": 38480, "loss": 1.4621, "lr": 6.54885654885655e-06, "epoch": 0.2624740124740125, "percentage": 1.31, "elapsed_time": "0:01:10", "remaining_time": "1:28:55", "throughput": 1349.34, "total_tokens": 95744}
+{"current_steps": 510, "total_steps": 38480, "loss": 1.1065, "lr": 6.613825363825364e-06, "epoch": 0.26507276507276506, "percentage": 1.33, "elapsed_time": "0:01:11", "remaining_time": "1:28:54", "throughput": 1349.58, "total_tokens": 96704}
+{"current_steps": 515, "total_steps": 38480, "loss": 1.057, "lr": 6.678794178794179e-06, "epoch": 0.26767151767151764, "percentage": 1.34, "elapsed_time": "0:01:12", "remaining_time": "1:28:53", "throughput": 1350.71, "total_tokens": 97728}
+{"current_steps": 520, "total_steps": 38480, "loss": 1.5113, "lr": 6.743762993762993e-06, "epoch": 0.2702702702702703, "percentage": 1.35, "elapsed_time": "0:01:13", "remaining_time": "1:28:52", "throughput": 1350.93, "total_tokens": 98688}
+{"current_steps": 525, "total_steps": 38480, "loss": 1.6482, "lr": 6.808731808731809e-06, "epoch": 0.27286902286902287, "percentage": 1.36, "elapsed_time": "0:01:13", "remaining_time": "1:28:51", "throughput": 1351.15, "total_tokens": 99648}
+{"current_steps": 530, "total_steps": 38480, "loss": 1.7921, "lr": 6.873700623700624e-06, "epoch": 0.27546777546777546, "percentage": 1.38, "elapsed_time": "0:01:14", "remaining_time": "1:28:51", "throughput": 1350.44, "total_tokens": 100544}
+{"current_steps": 535, "total_steps": 38480, "loss": 1.4071, "lr": 6.9386694386694386e-06, "epoch": 0.27806652806652804, "percentage": 1.39, "elapsed_time": "0:01:15", "remaining_time": "1:28:50", "throughput": 1350.18, "total_tokens": 101472}
+{"current_steps": 540, "total_steps": 38480, "loss": 1.3702, "lr": 7.003638253638254e-06, "epoch": 0.2806652806652807, "percentage": 1.4, "elapsed_time": "0:01:15", "remaining_time": "1:28:49", "throughput": 1349.99, "total_tokens": 102400}
+{"current_steps": 545, "total_steps": 38480, "loss": 1.1136, "lr": 7.0686070686070696e-06, "epoch": 0.28326403326403327, "percentage": 1.42, "elapsed_time": "0:01:16", "remaining_time": "1:28:48", "throughput": 1351.12, "total_tokens": 103424}
+{"current_steps": 550, "total_steps": 38480, "loss": 1.3721, "lr": 7.133575883575884e-06, "epoch": 0.28586278586278585, "percentage": 1.43, "elapsed_time": "0:01:17", "remaining_time": "1:28:46", "throughput": 1351.82, "total_tokens": 104416}
+{"current_steps": 555, "total_steps": 38480, "loss": 1.0324, "lr": 7.198544698544699e-06, "epoch": 0.28846153846153844, "percentage": 1.44, "elapsed_time": "0:01:17", "remaining_time": "1:28:45", "throughput": 1351.32, "total_tokens": 105312}
+{"current_steps": 560, "total_steps": 38480, "loss": 1.3428, "lr": 7.263513513513513e-06, "epoch": 0.2910602910602911, "percentage": 1.46, "elapsed_time": "0:01:18", "remaining_time": "1:28:44", "throughput": 1352.03, "total_tokens": 106304}
+{"current_steps": 565, "total_steps": 38480, "loss": 1.2205, "lr": 7.328482328482329e-06, "epoch": 0.29365904365904366, "percentage": 1.47, "elapsed_time": "0:01:19", "remaining_time": "1:28:42", "throughput": 1351.53, "total_tokens": 107200}
+{"current_steps": 570, "total_steps": 38480, "loss": 1.4006, "lr": 7.393451143451144e-06, "epoch": 0.29625779625779625, "percentage": 1.48, "elapsed_time": "0:01:20", "remaining_time": "1:28:41", "throughput": 1351.81, "total_tokens": 108160}
+{"current_steps": 575, "total_steps": 38480, "loss": 1.2645, "lr": 7.458419958419958e-06, "epoch": 0.29885654885654883, "percentage": 1.49, "elapsed_time": "0:01:20", "remaining_time": "1:28:40", "throughput": 1352.08, "total_tokens": 109120}
+{"current_steps": 580, "total_steps": 38480, "loss": 1.1994, "lr": 7.523388773388773e-06, "epoch": 0.30145530145530147, "percentage": 1.51, "elapsed_time": "0:01:21", "remaining_time": "1:28:38", "throughput": 1352.75, "total_tokens": 110112}
+{"current_steps": 585, "total_steps": 38480, "loss": 1.1014, "lr": 7.588357588357589e-06, "epoch": 0.30405405405405406, "percentage": 1.52, "elapsed_time": "0:01:22", "remaining_time": "1:28:37", "throughput": 1353.03, "total_tokens": 111072}
+{"current_steps": 590, "total_steps": 38480, "loss": 1.1502, "lr": 7.653326403326404e-06, "epoch": 0.30665280665280664, "percentage": 1.53, "elapsed_time": "0:01:22", "remaining_time": "1:28:36", "throughput": 1353.29, "total_tokens": 112032}
+{"current_steps": 595, "total_steps": 38480, "loss": 0.6955, "lr": 7.718295218295219e-06, "epoch": 0.3092515592515592, "percentage": 1.55, "elapsed_time": "0:01:23", "remaining_time": "1:28:35", "throughput": 1353.51, "total_tokens": 112992}
+{"current_steps": 600, "total_steps": 38480, "loss": 0.8044, "lr": 7.783264033264033e-06, "epoch": 0.31185031185031187, "percentage": 1.56, "elapsed_time": "0:01:24", "remaining_time": "1:28:34", "throughput": 1354.14, "total_tokens": 113984}
+{"current_steps": 605, "total_steps": 38480, "loss": 1.0427, "lr": 7.848232848232849e-06, "epoch": 0.31444906444906445, "percentage": 1.57, "elapsed_time": "0:01:24", "remaining_time": "1:28:32", "throughput": 1354.78, "total_tokens": 114976}
+{"current_steps": 610, "total_steps": 38480, "loss": 0.9465, "lr": 7.913201663201663e-06, "epoch": 0.31704781704781704, "percentage": 1.59, "elapsed_time": "0:01:25", "remaining_time": "1:28:31", "throughput": 1356.87, "total_tokens": 116096}
+{"current_steps": 615, "total_steps": 38480, "loss": 0.6735, "lr": 7.978170478170479e-06, "epoch": 0.3196465696465696, "percentage": 1.6, "elapsed_time": "0:01:26", "remaining_time": "1:28:30", "throughput": 1356.73, "total_tokens": 117024}
+{"current_steps": 620, "total_steps": 38480, "loss": 0.6211, "lr": 8.043139293139293e-06, "epoch": 0.32224532224532226, "percentage": 1.61, "elapsed_time": "0:01:26", "remaining_time": "1:28:29", "throughput": 1356.98, "total_tokens": 117984}
+{"current_steps": 625, "total_steps": 38480, "loss": 0.88, "lr": 8.108108108108109e-06, "epoch": 0.32484407484407485, "percentage": 1.62, "elapsed_time": "0:01:27", "remaining_time": "1:28:28", "throughput": 1356.86, "total_tokens": 118912}
+{"current_steps": 630, "total_steps": 38480, "loss": 0.5626, "lr": 8.173076923076923e-06, "epoch": 0.32744282744282743, "percentage": 1.64, "elapsed_time": "0:01:28", "remaining_time": "1:28:27", "throughput": 1356.64, "total_tokens": 119840}
+{"current_steps": 635, "total_steps": 38480, "loss": 0.8788, "lr": 8.238045738045738e-06, "epoch": 0.33004158004158, "percentage": 1.65, "elapsed_time": "0:01:29", "remaining_time": "1:28:26", "throughput": 1356.45, "total_tokens": 120768}
+{"current_steps": 640, "total_steps": 38480, "loss": 1.0893, "lr": 8.303014553014553e-06, "epoch": 0.33264033264033266, "percentage": 1.66, "elapsed_time": "0:01:29", "remaining_time": "1:28:25", "throughput": 1357.68, "total_tokens": 121824}
+{"current_steps": 645, "total_steps": 38480, "loss": 0.3269, "lr": 8.36798336798337e-06, "epoch": 0.33523908523908524, "percentage": 1.68, "elapsed_time": "0:01:30", "remaining_time": "1:28:24", "throughput": 1357.09, "total_tokens": 122720}
+{"current_steps": 650, "total_steps": 38480, "loss": 0.7756, "lr": 8.432952182952184e-06, "epoch": 0.33783783783783783, "percentage": 1.69, "elapsed_time": "0:01:31", "remaining_time": "1:28:23", "throughput": 1356.95, "total_tokens": 123648}
+{"current_steps": 655, "total_steps": 38480, "loss": 0.4023, "lr": 8.497920997920998e-06, "epoch": 0.3404365904365904, "percentage": 1.7, "elapsed_time": "0:01:31", "remaining_time": "1:28:22", "throughput": 1357.52, "total_tokens": 124640}
+{"current_steps": 660, "total_steps": 38480, "loss": 1.0679, "lr": 8.562889812889812e-06, "epoch": 0.34303534303534305, "percentage": 1.72, "elapsed_time": "0:01:32", "remaining_time": "1:28:21", "throughput": 1357.64, "total_tokens": 125600}
+{"current_steps": 665, "total_steps": 38480, "loss": 0.5039, "lr": 8.627858627858628e-06, "epoch": 0.34563409563409564, "percentage": 1.73, "elapsed_time": "0:01:33", "remaining_time": "1:28:20", "throughput": 1357.42, "total_tokens": 126528}
+{"current_steps": 670, "total_steps": 38480, "loss": 0.6153, "lr": 8.692827442827444e-06, "epoch": 0.3482328482328482, "percentage": 1.74, "elapsed_time": "0:01:33", "remaining_time": "1:28:19", "throughput": 1357.93, "total_tokens": 127520}
+{"current_steps": 675, "total_steps": 38480, "loss": 0.3976, "lr": 8.757796257796258e-06, "epoch": 0.3508316008316008, "percentage": 1.75, "elapsed_time": "0:01:34", "remaining_time": "1:28:18", "throughput": 1357.45, "total_tokens": 128416}
+{"current_steps": 680, "total_steps": 38480, "loss": 0.271, "lr": 8.822765072765072e-06, "epoch": 0.35343035343035345, "percentage": 1.77, "elapsed_time": "0:01:35", "remaining_time": "1:28:17", "throughput": 1358.3, "total_tokens": 129440}
+{"current_steps": 685, "total_steps": 38480, "loss": 0.9601, "lr": 8.887733887733888e-06, "epoch": 0.35602910602910603, "percentage": 1.78, "elapsed_time": "0:01:35", "remaining_time": "1:28:16", "throughput": 1358.79, "total_tokens": 130432}
+{"current_steps": 690, "total_steps": 38480, "loss": 0.6837, "lr": 8.952702702702702e-06, "epoch": 0.3586278586278586, "percentage": 1.79, "elapsed_time": "0:01:36", "remaining_time": "1:28:15", "throughput": 1358.92, "total_tokens": 131392}
+{"current_steps": 695, "total_steps": 38480, "loss": 0.4798, "lr": 9.017671517671518e-06, "epoch": 0.3612266112266112, "percentage": 1.81, "elapsed_time": "0:01:37", "remaining_time": "1:28:14", "throughput": 1358.72, "total_tokens": 132320}
+{"current_steps": 700, "total_steps": 38480, "loss": 0.5671, "lr": 9.082640332640333e-06, "epoch": 0.36382536382536385, "percentage": 1.82, "elapsed_time": "0:01:38", "remaining_time": "1:28:13", "throughput": 1357.62, "total_tokens": 133152}
+{"current_steps": 705, "total_steps": 38480, "loss": 0.3383, "lr": 9.147609147609149e-06, "epoch": 0.36642411642411643, "percentage": 1.83, "elapsed_time": "0:01:38", "remaining_time": "1:28:12", "throughput": 1358.15, "total_tokens": 134144}
+{"current_steps": 710, "total_steps": 38480, "loss": 0.6047, "lr": 9.212577962577963e-06, "epoch": 0.369022869022869, "percentage": 1.85, "elapsed_time": "0:01:39", "remaining_time": "1:28:11", "throughput": 1357.4, "total_tokens": 135008}
+{"current_steps": 715, "total_steps": 38480, "loss": 0.6505, "lr": 9.277546777546777e-06, "epoch": 0.3716216216216216, "percentage": 1.86, "elapsed_time": "0:01:40", "remaining_time": "1:28:10", "throughput": 1357.25, "total_tokens": 135936}
+{"current_steps": 720, "total_steps": 38480, "loss": 0.6104, "lr": 9.342515592515593e-06, "epoch": 0.37422037422037424, "percentage": 1.87, "elapsed_time": "0:01:40", "remaining_time": "1:28:09", "throughput": 1357.71, "total_tokens": 136928}
+{"current_steps": 725, "total_steps": 38480, "loss": 0.2644, "lr": 9.407484407484409e-06, "epoch": 0.3768191268191268, "percentage": 1.88, "elapsed_time": "0:01:41", "remaining_time": "1:28:08", "throughput": 1357.85, "total_tokens": 137888}
+{"current_steps": 730, "total_steps": 38480, "loss": 0.6761, "lr": 9.472453222453223e-06, "epoch": 0.3794178794178794, "percentage": 1.9, "elapsed_time": "0:01:42", "remaining_time": "1:28:07", "throughput": 1358.33, "total_tokens": 138880}
+{"current_steps": 735, "total_steps": 38480, "loss": 0.5843, "lr": 9.537422037422037e-06, "epoch": 0.382016632016632, "percentage": 1.91, "elapsed_time": "0:01:42", "remaining_time": "1:28:06", "throughput": 1358.74, "total_tokens": 139872}
+{"current_steps": 740, "total_steps": 38480, "loss": 0.2, "lr": 9.602390852390852e-06, "epoch": 0.38461538461538464, "percentage": 1.92, "elapsed_time": "0:01:43", "remaining_time": "1:28:05", "throughput": 1358.26, "total_tokens": 140768}
+{"current_steps": 745, "total_steps": 38480, "loss": 1.0062, "lr": 9.667359667359669e-06, "epoch": 0.3872141372141372, "percentage": 1.94, "elapsed_time": "0:01:44", "remaining_time": "1:28:04", "throughput": 1358.69, "total_tokens": 141760}
+{"current_steps": 750, "total_steps": 38480, "loss": 0.3925, "lr": 9.732328482328483e-06, "epoch": 0.3898128898128898, "percentage": 1.95, "elapsed_time": "0:01:45", "remaining_time": "1:28:03", "throughput": 1358.47, "total_tokens": 142688}
+{"current_steps": 755, "total_steps": 38480, "loss": 0.216, "lr": 9.797297297297298e-06, "epoch": 0.3924116424116424, "percentage": 1.96, "elapsed_time": "0:01:45", "remaining_time": "1:28:03", "throughput": 1358.82, "total_tokens": 143680}
+{"current_steps": 760, "total_steps": 38480, "loss": 0.5414, "lr": 9.862266112266112e-06, "epoch": 0.39501039501039503, "percentage": 1.98, "elapsed_time": "0:01:46", "remaining_time": "1:28:02", "throughput": 1359.29, "total_tokens": 144672}
+{"current_steps": 765, "total_steps": 38480, "loss": 0.2889, "lr": 9.927234927234928e-06, "epoch": 0.3976091476091476, "percentage": 1.99, "elapsed_time": "0:01:47", "remaining_time": "1:28:01", "throughput": 1358.55, "total_tokens": 145536}
+{"current_steps": 770, "total_steps": 38480, "loss": 0.4749, "lr": 9.992203742203744e-06, "epoch": 0.4002079002079002, "percentage": 2.0, "elapsed_time": "0:01:47", "remaining_time": "1:28:00", "throughput": 1358.38, "total_tokens": 146464}
+{"current_steps": 775, "total_steps": 38480, "loss": 0.4759, "lr": 1.0057172557172558e-05, "epoch": 0.4028066528066528, "percentage": 2.01, "elapsed_time": "0:01:48", "remaining_time": "1:27:59", "throughput": 1359.38, "total_tokens": 147520}
+{"current_steps": 780, "total_steps": 38480, "loss": 0.4036, "lr": 1.0122141372141372e-05, "epoch": 0.40540540540540543, "percentage": 2.03, "elapsed_time": "0:01:49", "remaining_time": "1:27:58", "throughput": 1359.49, "total_tokens": 148480}
+{"current_steps": 785, "total_steps": 38480, "loss": 0.1976, "lr": 1.0187110187110188e-05, "epoch": 0.408004158004158, "percentage": 2.04, "elapsed_time": "0:01:49", "remaining_time": "1:27:57", "throughput": 1359.64, "total_tokens": 149440}
+{"current_steps": 790, "total_steps": 38480, "loss": 0.4456, "lr": 1.0252079002079002e-05, "epoch": 0.4106029106029106, "percentage": 2.05, "elapsed_time": "0:01:50", "remaining_time": "1:27:56", "throughput": 1359.21, "total_tokens": 150336}
+{"current_steps": 795, "total_steps": 38480, "loss": 0.1852, "lr": 1.0317047817047818e-05, "epoch": 0.4132016632016632, "percentage": 2.07, "elapsed_time": "0:01:51", "remaining_time": "1:27:55", "throughput": 1358.78, "total_tokens": 151232}
+{"current_steps": 800, "total_steps": 38480, "loss": 0.4057, "lr": 1.0382016632016632e-05, "epoch": 0.4158004158004158, "percentage": 2.08, "elapsed_time": "0:01:51", "remaining_time": "1:27:54", "throughput": 1358.65, "total_tokens": 152160}
+{"current_steps": 805, "total_steps": 38480, "loss": 0.314, "lr": 1.0446985446985448e-05, "epoch": 0.4183991683991684, "percentage": 2.09, "elapsed_time": "0:01:52", "remaining_time": "1:27:53", "throughput": 1358.79, "total_tokens": 153120}
+{"current_steps": 810, "total_steps": 38480, "loss": 0.4367, "lr": 1.0511954261954263e-05, "epoch": 0.420997920997921, "percentage": 2.1, "elapsed_time": "0:01:53", "remaining_time": "1:27:52", "throughput": 1359.23, "total_tokens": 154112}
+{"current_steps": 815, "total_steps": 38480, "loss": 0.453, "lr": 1.0576923076923077e-05, "epoch": 0.4235966735966736, "percentage": 2.12, "elapsed_time": "0:01:54", "remaining_time": "1:27:51", "throughput": 1358.83, "total_tokens": 155008}
+{"current_steps": 820, "total_steps": 38480, "loss": 0.1357, "lr": 1.0641891891891893e-05, "epoch": 0.4261954261954262, "percentage": 2.13, "elapsed_time": "0:01:54", "remaining_time": "1:27:51", "throughput": 1359.23, "total_tokens": 156000}
+{"current_steps": 825, "total_steps": 38480, "loss": 0.4098, "lr": 1.0706860706860709e-05, "epoch": 0.4287941787941788, "percentage": 2.14, "elapsed_time": "0:01:55", "remaining_time": "1:27:50", "throughput": 1359.09, "total_tokens": 156928}
+{"current_steps": 830, "total_steps": 38480, "loss": 0.4344, "lr": 1.0771829521829523e-05, "epoch": 0.4313929313929314, "percentage": 2.16, "elapsed_time": "0:01:56", "remaining_time": "1:27:49", "throughput": 1359.24, "total_tokens": 157888}
+{"current_steps": 835, "total_steps": 38480, "loss": 0.4487, "lr": 1.0836798336798337e-05, "epoch": 0.433991683991684, "percentage": 2.17, "elapsed_time": "0:01:56", "remaining_time": "1:27:48", "throughput": 1359.39, "total_tokens": 158848}
+{"current_steps": 840, "total_steps": 38480, "loss": 0.4532, "lr": 1.0901767151767151e-05, "epoch": 0.4365904365904366, "percentage": 2.18, "elapsed_time": "0:01:57", "remaining_time": "1:27:47", "throughput": 1359.25, "total_tokens": 159776}
+{"current_steps": 845, "total_steps": 38480, "loss": 0.4106, "lr": 1.0966735966735967e-05, "epoch": 0.4391891891891892, "percentage": 2.2, "elapsed_time": "0:01:58", "remaining_time": "1:27:46", "throughput": 1359.92, "total_tokens": 160800}
+{"current_steps": 850, "total_steps": 38480, "loss": 0.4015, "lr": 1.1031704781704783e-05, "epoch": 0.4417879417879418, "percentage": 2.21, "elapsed_time": "0:01:58", "remaining_time": "1:27:45", "throughput": 1360.34, "total_tokens": 161792}
+{"current_steps": 855, "total_steps": 38480, "loss": 0.2682, "lr": 1.1096673596673597e-05, "epoch": 0.44438669438669437, "percentage": 2.22, "elapsed_time": "0:01:59", "remaining_time": "1:27:44", "throughput": 1360.21, "total_tokens": 162720}
+{"current_steps": 860, "total_steps": 38480, "loss": 0.5684, "lr": 1.1161642411642412e-05, "epoch": 0.446985446985447, "percentage": 2.23, "elapsed_time": "0:02:00", "remaining_time": "1:27:43", "throughput": 1360.24, "total_tokens": 163680}
+{"current_steps": 865, "total_steps": 38480, "loss": 0.5395, "lr": 1.1226611226611228e-05, "epoch": 0.4495841995841996, "percentage": 2.25, "elapsed_time": "0:02:01", "remaining_time": "1:27:43", "throughput": 1360.6, "total_tokens": 164672}
+{"current_steps": 870, "total_steps": 38480, "loss": 0.2689, "lr": 1.1291580041580042e-05, "epoch": 0.4521829521829522, "percentage": 2.26, "elapsed_time": "0:02:01", "remaining_time": "1:27:41", "throughput": 1360.22, "total_tokens": 165568}
+{"current_steps": 875, "total_steps": 38480, "loss": 0.3476, "lr": 1.1356548856548858e-05, "epoch": 0.45478170478170477, "percentage": 2.27, "elapsed_time": "0:02:02", "remaining_time": "1:27:41", "throughput": 1360.06, "total_tokens": 166496}
+{"current_steps": 880, "total_steps": 38480, "loss": 0.4811, "lr": 1.1421517671517672e-05, "epoch": 0.4573804573804574, "percentage": 2.29, "elapsed_time": "0:02:03", "remaining_time": "1:27:40", "throughput": 1360.16, "total_tokens": 167456}
+{"current_steps": 885, "total_steps": 38480, "loss": 0.343, "lr": 1.1486486486486488e-05, "epoch": 0.45997920997921, "percentage": 2.3, "elapsed_time": "0:02:03", "remaining_time": "1:27:39", "throughput": 1360.28, "total_tokens": 168416}
+{"current_steps": 890, "total_steps": 38480, "loss": 0.3567, "lr": 1.1551455301455302e-05, "epoch": 0.4625779625779626, "percentage": 2.31, "elapsed_time": "0:02:04", "remaining_time": "1:27:38", "throughput": 1359.9, "total_tokens": 169312}
+{"current_steps": 895, "total_steps": 38480, "loss": 0.4487, "lr": 1.1616424116424116e-05, "epoch": 0.46517671517671516, "percentage": 2.33, "elapsed_time": "0:02:05", "remaining_time": "1:27:37", "throughput": 1360.0, "total_tokens": 170272}
+{"current_steps": 900, "total_steps": 38480, "loss": 0.3058, "lr": 1.1681392931392932e-05, "epoch": 0.4677754677754678, "percentage": 2.34, "elapsed_time": "0:02:05", "remaining_time": "1:27:36", "throughput": 1359.32, "total_tokens": 171136}
+{"current_steps": 905, "total_steps": 38480, "loss": 0.1231, "lr": 1.1746361746361748e-05, "epoch": 0.4703742203742204, "percentage": 2.35, "elapsed_time": "0:02:06", "remaining_time": "1:27:36", "throughput": 1358.92, "total_tokens": 172032}
+{"current_steps": 910, "total_steps": 38480, "loss": 0.4516, "lr": 1.1811330561330562e-05, "epoch": 0.47297297297297297, "percentage": 2.36, "elapsed_time": "0:02:07", "remaining_time": "1:27:35", "throughput": 1359.01, "total_tokens": 172992}
+{"current_steps": 915, "total_steps": 38480, "loss": 0.4094, "lr": 1.1876299376299377e-05, "epoch": 0.47557172557172556, "percentage": 2.38, "elapsed_time": "0:02:07", "remaining_time": "1:27:34", "throughput": 1359.1, "total_tokens": 173952}
+{"current_steps": 920, "total_steps": 38480, "loss": 0.4099, "lr": 1.194126819126819e-05, "epoch": 0.4781704781704782, "percentage": 2.39, "elapsed_time": "0:02:08", "remaining_time": "1:27:33", "throughput": 1358.69, "total_tokens": 174848}
+{"current_steps": 925, "total_steps": 38480, "loss": 0.4286, "lr": 1.2006237006237007e-05, "epoch": 0.4807692307692308, "percentage": 2.4, "elapsed_time": "0:02:09", "remaining_time": "1:27:32", "throughput": 1358.57, "total_tokens": 175776}
+{"current_steps": 930, "total_steps": 38480, "loss": 0.3136, "lr": 1.2071205821205823e-05, "epoch": 0.48336798336798337, "percentage": 2.42, "elapsed_time": "0:02:10", "remaining_time": "1:27:32", "throughput": 1359.66, "total_tokens": 176864}
+{"current_steps": 935, "total_steps": 38480, "loss": 0.3767, "lr": 1.2136174636174637e-05, "epoch": 0.48596673596673595, "percentage": 2.43, "elapsed_time": "0:02:10", "remaining_time": "1:27:31", "throughput": 1360.02, "total_tokens": 177856}
+{"current_steps": 940, "total_steps": 38480, "loss": 0.3064, "lr": 1.2201143451143451e-05, "epoch": 0.4885654885654886, "percentage": 2.44, "elapsed_time": "0:02:11", "remaining_time": "1:27:30", "throughput": 1360.34, "total_tokens": 178848}
+{"current_steps": 945, "total_steps": 38480, "loss": 0.5884, "lr": 1.2266112266112267e-05, "epoch": 0.4911642411642412, "percentage": 2.46, "elapsed_time": "0:02:12", "remaining_time": "1:27:29", "throughput": 1360.9, "total_tokens": 179872}
+{"current_steps": 950, "total_steps": 38480, "loss": 0.4918, "lr": 1.2331081081081081e-05, "epoch": 0.49376299376299376, "percentage": 2.47, "elapsed_time": "0:02:12", "remaining_time": "1:27:28", "throughput": 1360.03, "total_tokens": 180704}
+{"current_steps": 955, "total_steps": 38480, "loss": 0.1717, "lr": 1.2396049896049897e-05, "epoch": 0.49636174636174635, "percentage": 2.48, "elapsed_time": "0:02:13", "remaining_time": "1:27:28", "throughput": 1359.4, "total_tokens": 181568}
+{"current_steps": 960, "total_steps": 38480, "loss": 0.3197, "lr": 1.2461018711018711e-05, "epoch": 0.498960498960499, "percentage": 2.49, "elapsed_time": "0:02:14", "remaining_time": "1:27:27", "throughput": 1359.96, "total_tokens": 182592}
+{"current_steps": 965, "total_steps": 38480, "loss": 0.2191, "lr": 1.2525987525987526e-05, "epoch": 0.5015592515592515, "percentage": 2.51, "elapsed_time": "0:02:14", "remaining_time": "1:27:26", "throughput": 1360.5, "total_tokens": 183616}
+{"current_steps": 970, "total_steps": 38480, "loss": 0.2669, "lr": 1.2590956340956342e-05, "epoch": 0.5041580041580042, "percentage": 2.52, "elapsed_time": "0:02:15", "remaining_time": "1:27:26", "throughput": 1360.72, "total_tokens": 184608}
+{"current_steps": 975, "total_steps": 38480, "loss": 0.3562, "lr": 1.2655925155925156e-05, "epoch": 0.5067567567567568, "percentage": 2.53, "elapsed_time": "0:02:16", "remaining_time": "1:27:25", "throughput": 1360.56, "total_tokens": 185536}
+{"current_steps": 980, "total_steps": 38480, "loss": 0.3366, "lr": 1.2720893970893972e-05, "epoch": 0.5093555093555093, "percentage": 2.55, "elapsed_time": "0:02:17", "remaining_time": "1:27:24", "throughput": 1361.1, "total_tokens": 186560}
+{"current_steps": 985, "total_steps": 38480, "loss": 0.4887, "lr": 1.2785862785862788e-05, "epoch": 0.511954261954262, "percentage": 2.56, "elapsed_time": "0:02:17", "remaining_time": "1:27:24", "throughput": 1361.4, "total_tokens": 187552}
+{"current_steps": 990, "total_steps": 38480, "loss": 0.2911, "lr": 1.28508316008316e-05, "epoch": 0.5145530145530145, "percentage": 2.57, "elapsed_time": "0:02:18", "remaining_time": "1:27:23", "throughput": 1361.95, "total_tokens": 188576}
+{"current_steps": 995, "total_steps": 38480, "loss": 0.3361, "lr": 1.2915800415800416e-05, "epoch": 0.5171517671517671, "percentage": 2.59, "elapsed_time": "0:02:19", "remaining_time": "1:27:22", "throughput": 1361.59, "total_tokens": 189472}
+{"current_steps": 1000, "total_steps": 38480, "loss": 0.3709, "lr": 1.2980769230769232e-05, "epoch": 0.5197505197505198, "percentage": 2.6, "elapsed_time": "0:02:19", "remaining_time": "1:27:21", "throughput": 1361.22, "total_tokens": 190368}
+{"current_steps": 1005, "total_steps": 38480, "loss": 0.0203, "lr": 1.3045738045738046e-05, "epoch": 0.5223492723492723, "percentage": 2.61, "elapsed_time": "0:02:20", "remaining_time": "1:27:20", "throughput": 1361.06, "total_tokens": 191296}
+{"current_steps": 1010, "total_steps": 38480, "loss": 0.6673, "lr": 1.3110706860706862e-05, "epoch": 0.524948024948025, "percentage": 2.62, "elapsed_time": "0:02:21", "remaining_time": "1:27:20", "throughput": 1360.69, "total_tokens": 192192}
+{"current_steps": 1015, "total_steps": 38480, "loss": 0.3606, "lr": 1.3175675675675675e-05, "epoch": 0.5275467775467776, "percentage": 2.64, "elapsed_time": "0:02:21", "remaining_time": "1:27:19", "throughput": 1360.77, "total_tokens": 193152}
+{"current_steps": 1020, "total_steps": 38480, "loss": 0.729, "lr": 1.324064449064449e-05, "epoch": 0.5301455301455301, "percentage": 2.65, "elapsed_time": "0:02:22", "remaining_time": "1:27:18", "throughput": 1360.85, "total_tokens": 194112}
+{"current_steps": 1025, "total_steps": 38480, "loss": 0.2439, "lr": 1.3305613305613307e-05, "epoch": 0.5327442827442828, "percentage": 2.66, "elapsed_time": "0:02:23", "remaining_time": "1:27:17", "throughput": 1360.94, "total_tokens": 195072}
+{"current_steps": 1030, "total_steps": 38480, "loss": 0.2235, "lr": 1.337058212058212e-05, "epoch": 0.5353430353430353, "percentage": 2.68, "elapsed_time": "0:02:24", "remaining_time": "1:27:16", "throughput": 1360.39, "total_tokens": 195936}
+{"current_steps": 1035, "total_steps": 38480, "loss": 0.4322, "lr": 1.3435550935550937e-05, "epoch": 0.5379417879417879, "percentage": 2.69, "elapsed_time": "0:02:24", "remaining_time": "1:27:15", "throughput": 1360.28, "total_tokens": 196864}
+{"current_steps": 1040, "total_steps": 38480, "loss": 0.2937, "lr": 1.3500519750519753e-05, "epoch": 0.5405405405405406, "percentage": 2.7, "elapsed_time": "0:02:25", "remaining_time": "1:27:14", "throughput": 1360.84, "total_tokens": 197888}
+{"current_steps": 1045, "total_steps": 38480, "loss": 0.1359, "lr": 1.3565488565488565e-05, "epoch": 0.5431392931392931, "percentage": 2.72, "elapsed_time": "0:02:26", "remaining_time": "1:27:14", "throughput": 1360.75, "total_tokens": 198816}
+{"current_steps": 1050, "total_steps": 38480, "loss": 0.4252, "lr": 1.3630457380457381e-05, "epoch": 0.5457380457380457, "percentage": 2.73, "elapsed_time": "0:02:26", "remaining_time": "1:27:13", "throughput": 1360.6, "total_tokens": 199744}
+{"current_steps": 1055, "total_steps": 38480, "loss": 0.2602, "lr": 1.3695426195426195e-05, "epoch": 0.5483367983367984, "percentage": 2.74, "elapsed_time": "0:02:27", "remaining_time": "1:27:12", "throughput": 1360.67, "total_tokens": 200704}
+{"current_steps": 1060, "total_steps": 38480, "loss": 0.115, "lr": 1.3760395010395011e-05, "epoch": 0.5509355509355509, "percentage": 2.75, "elapsed_time": "0:02:28", "remaining_time": "1:27:11", "throughput": 1360.75, "total_tokens": 201664}
+{"current_steps": 1065, "total_steps": 38480, "loss": 0.4572, "lr": 1.3825363825363827e-05, "epoch": 0.5535343035343036, "percentage": 2.77, "elapsed_time": "0:02:28", "remaining_time": "1:27:11", "throughput": 1360.81, "total_tokens": 202624}
+{"current_steps": 1070, "total_steps": 38480, "loss": 0.0809, "lr": 1.389033264033264e-05, "epoch": 0.5561330561330561, "percentage": 2.78, "elapsed_time": "0:02:29", "remaining_time": "1:27:10", "throughput": 1360.46, "total_tokens": 203520}
+{"current_steps": 1075, "total_steps": 38480, "loss": 0.2359, "lr": 1.3955301455301456e-05, "epoch": 0.5587318087318087, "percentage": 2.79, "elapsed_time": "0:02:30", "remaining_time": "1:27:09", "throughput": 1360.68, "total_tokens": 204512}
+{"current_steps": 1080, "total_steps": 38480, "loss": 0.2572, "lr": 1.4020270270270271e-05, "epoch": 0.5613305613305614, "percentage": 2.81, "elapsed_time": "0:02:31", "remaining_time": "1:27:09", "throughput": 1360.74, "total_tokens": 205472}
+{"current_steps": 1085, "total_steps": 38480, "loss": 0.4012, "lr": 1.4085239085239086e-05, "epoch": 0.5639293139293139, "percentage": 2.82, "elapsed_time": "0:02:31", "remaining_time": "1:27:08", "throughput": 1361.44, "total_tokens": 206528}
+{"current_steps": 1090, "total_steps": 38480, "loss": 0.2503, "lr": 1.4150207900207902e-05, "epoch": 0.5665280665280665, "percentage": 2.83, "elapsed_time": "0:02:32", "remaining_time": "1:27:07", "throughput": 1361.1, "total_tokens": 207424}
+{"current_steps": 1095, "total_steps": 38480, "loss": 0.081, "lr": 1.4215176715176714e-05, "epoch": 0.5691268191268192, "percentage": 2.85, "elapsed_time": "0:02:33", "remaining_time": "1:27:06", "throughput": 1361.4, "total_tokens": 208416}
+{"current_steps": 1100, "total_steps": 38480, "loss": 0.2873, "lr": 1.428014553014553e-05, "epoch": 0.5717255717255717, "percentage": 2.86, "elapsed_time": "0:02:33", "remaining_time": "1:27:05", "throughput": 1361.71, "total_tokens": 209408}
+{"current_steps": 1105, "total_steps": 38480, "loss": 0.3881, "lr": 1.4345114345114346e-05, "epoch": 0.5743243243243243, "percentage": 2.87, "elapsed_time": "0:02:34", "remaining_time": "1:27:04", "throughput": 1362.23, "total_tokens": 210432}
+{"current_steps": 1110, "total_steps": 38480, "loss": 0.0864, "lr": 1.441008316008316e-05, "epoch": 0.5769230769230769, "percentage": 2.88, "elapsed_time": "0:02:35", "remaining_time": "1:27:04", "throughput": 1362.12, "total_tokens": 211360}
+{"current_steps": 1115, "total_steps": 38480, "loss": 0.3045, "lr": 1.4475051975051976e-05, "epoch": 0.5795218295218295, "percentage": 2.9, "elapsed_time": "0:02:35", "remaining_time": "1:27:03", "throughput": 1362.02, "total_tokens": 212288}
+{"current_steps": 1120, "total_steps": 38480, "loss": 0.33, "lr": 1.4540020790020792e-05, "epoch": 0.5821205821205822, "percentage": 2.91, "elapsed_time": "0:02:36", "remaining_time": "1:27:02", "throughput": 1361.92, "total_tokens": 213216}
+{"current_steps": 1125, "total_steps": 38480, "loss": 0.31, "lr": 1.4604989604989605e-05, "epoch": 0.5847193347193347, "percentage": 2.92, "elapsed_time": "0:02:37", "remaining_time": "1:27:01", "throughput": 1362.23, "total_tokens": 214208}
+{"current_steps": 1130, "total_steps": 38480, "loss": 0.3037, "lr": 1.466995841995842e-05, "epoch": 0.5873180873180873, "percentage": 2.94, "elapsed_time": "0:02:37", "remaining_time": "1:27:00", "throughput": 1362.74, "total_tokens": 215232}
+{"current_steps": 1135, "total_steps": 38480, "loss": 0.3705, "lr": 1.4734927234927235e-05, "epoch": 0.58991683991684, "percentage": 2.95, "elapsed_time": "0:02:38", "remaining_time": "1:26:59", "throughput": 1362.25, "total_tokens": 216096}
+{"current_steps": 1140, "total_steps": 38480, "loss": 0.2918, "lr": 1.479989604989605e-05, "epoch": 0.5925155925155925, "percentage": 2.96, "elapsed_time": "0:02:39", "remaining_time": "1:26:58", "throughput": 1362.13, "total_tokens": 217024}
+{"current_steps": 1145, "total_steps": 38480, "loss": 0.1559, "lr": 1.4864864864864867e-05, "epoch": 0.5951143451143451, "percentage": 2.98, "elapsed_time": "0:02:40", "remaining_time": "1:26:57", "throughput": 1361.81, "total_tokens": 217920}
+{"current_steps": 1150, "total_steps": 38480, "loss": 0.5104, "lr": 1.4929833679833679e-05, "epoch": 0.5977130977130977, "percentage": 2.99, "elapsed_time": "0:02:40", "remaining_time": "1:26:56", "throughput": 1361.52, "total_tokens": 218816}
+{"current_steps": 1155, "total_steps": 38480, "loss": 0.3781, "lr": 1.4994802494802495e-05, "epoch": 0.6003118503118503, "percentage": 3.0, "elapsed_time": "0:02:41", "remaining_time": "1:26:56", "throughput": 1361.62, "total_tokens": 219776}
+{"current_steps": 1160, "total_steps": 38480, "loss": 0.2502, "lr": 1.5059771309771311e-05, "epoch": 0.6029106029106029, "percentage": 3.01, "elapsed_time": "0:02:42", "remaining_time": "1:26:55", "throughput": 1361.49, "total_tokens": 220704}
+{"current_steps": 1165, "total_steps": 38480, "loss": 0.2797, "lr": 1.5124740124740125e-05, "epoch": 0.6055093555093555, "percentage": 3.03, "elapsed_time": "0:02:42", "remaining_time": "1:26:54", "throughput": 1362.38, "total_tokens": 221792}
+{"current_steps": 1170, "total_steps": 38480, "loss": 0.7419, "lr": 1.5189708939708941e-05, "epoch": 0.6081081081081081, "percentage": 3.04, "elapsed_time": "0:02:43", "remaining_time": "1:26:53", "throughput": 1362.67, "total_tokens": 222784}
+{"current_steps": 1175, "total_steps": 38480, "loss": 0.392, "lr": 1.5254677754677754e-05, "epoch": 0.6107068607068608, "percentage": 3.05, "elapsed_time": "0:02:44", "remaining_time": "1:26:52", "throughput": 1362.76, "total_tokens": 223744}
+{"current_steps": 1180, "total_steps": 38480, "loss": 0.1593, "lr": 1.531964656964657e-05, "epoch": 0.6133056133056133, "percentage": 3.07, "elapsed_time": "0:02:44", "remaining_time": "1:26:51", "throughput": 1363.04, "total_tokens": 224736}
+{"current_steps": 1185, "total_steps": 38480, "loss": 0.2469, "lr": 1.5384615384615387e-05, "epoch": 0.6159043659043659, "percentage": 3.08, "elapsed_time": "0:02:45", "remaining_time": "1:26:51", "throughput": 1363.1, "total_tokens": 225696}
+{"current_steps": 1190, "total_steps": 38480, "loss": 0.3126, "lr": 1.54495841995842e-05, "epoch": 0.6185031185031185, "percentage": 3.09, "elapsed_time": "0:02:46", "remaining_time": "1:26:50", "throughput": 1363.32, "total_tokens": 226688}
+{"current_steps": 1195, "total_steps": 38480, "loss": 0.1982, "lr": 1.5514553014553016e-05, "epoch": 0.6211018711018711, "percentage": 3.11, "elapsed_time": "0:02:46", "remaining_time": "1:26:49", "throughput": 1363.54, "total_tokens": 227680}
+{"current_steps": 1200, "total_steps": 38480, "loss": 0.2087, "lr": 1.557952182952183e-05, "epoch": 0.6237006237006237, "percentage": 3.12, "elapsed_time": "0:02:47", "remaining_time": "1:26:48", "throughput": 1363.24, "total_tokens": 228576}
+{"current_steps": 1205, "total_steps": 38480, "loss": 0.1441, "lr": 1.5644490644490644e-05, "epoch": 0.6262993762993763, "percentage": 3.13, "elapsed_time": "0:02:48", "remaining_time": "1:26:48", "throughput": 1363.14, "total_tokens": 229504}
+{"current_steps": 1210, "total_steps": 38480, "loss": 0.124, "lr": 1.570945945945946e-05, "epoch": 0.6288981288981289, "percentage": 3.14, "elapsed_time": "0:02:49", "remaining_time": "1:26:47", "throughput": 1363.23, "total_tokens": 230464}
+{"current_steps": 1215, "total_steps": 38480, "loss": 0.1808, "lr": 1.5774428274428273e-05, "epoch": 0.6314968814968815, "percentage": 3.16, "elapsed_time": "0:02:49", "remaining_time": "1:26:46", "throughput": 1363.49, "total_tokens": 231456}
+{"current_steps": 1220, "total_steps": 38480, "loss": 0.3585, "lr": 1.583939708939709e-05, "epoch": 0.6340956340956341, "percentage": 3.17, "elapsed_time": "0:02:50", "remaining_time": "1:26:45", "throughput": 1362.98, "total_tokens": 232320}
+{"current_steps": 1225, "total_steps": 38480, "loss": 0.2376, "lr": 1.5904365904365904e-05, "epoch": 0.6366943866943867, "percentage": 3.18, "elapsed_time": "0:02:51", "remaining_time": "1:26:44", "throughput": 1362.51, "total_tokens": 233184}
+{"current_steps": 1230, "total_steps": 38480, "loss": 0.338, "lr": 1.596933471933472e-05, "epoch": 0.6392931392931392, "percentage": 3.2, "elapsed_time": "0:02:51", "remaining_time": "1:26:44", "throughput": 1362.4, "total_tokens": 234112}
+{"current_steps": 1235, "total_steps": 38480, "loss": 0.3006, "lr": 1.6034303534303536e-05, "epoch": 0.6418918918918919, "percentage": 3.21, "elapsed_time": "0:02:52", "remaining_time": "1:26:43", "throughput": 1362.28, "total_tokens": 235040}
+{"current_steps": 1240, "total_steps": 38480, "loss": 0.2039, "lr": 1.6099272349272352e-05, "epoch": 0.6444906444906445, "percentage": 3.22, "elapsed_time": "0:02:53", "remaining_time": "1:26:42", "throughput": 1362.52, "total_tokens": 236032}
+{"current_steps": 1245, "total_steps": 38480, "loss": 0.2783, "lr": 1.6164241164241165e-05, "epoch": 0.6470893970893971, "percentage": 3.24, "elapsed_time": "0:02:53", "remaining_time": "1:26:41", "throughput": 1362.59, "total_tokens": 236992}
+{"current_steps": 1250, "total_steps": 38480, "loss": 0.0536, "lr": 1.622920997920998e-05, "epoch": 0.6496881496881497, "percentage": 3.25, "elapsed_time": "0:02:54", "remaining_time": "1:26:40", "throughput": 1363.04, "total_tokens": 238016}
+{"current_steps": 1255, "total_steps": 38480, "loss": 0.3023, "lr": 1.6294178794178793e-05, "epoch": 0.6522869022869023, "percentage": 3.26, "elapsed_time": "0:02:55", "remaining_time": "1:26:40", "throughput": 1362.94, "total_tokens": 238944}
+{"current_steps": 1260, "total_steps": 38480, "loss": 0.298, "lr": 1.635914760914761e-05, "epoch": 0.6548856548856549, "percentage": 3.27, "elapsed_time": "0:02:56", "remaining_time": "1:26:39", "throughput": 1363.02, "total_tokens": 239904}
+{"current_steps": 1265, "total_steps": 38480, "loss": 0.2526, "lr": 1.6424116424116425e-05, "epoch": 0.6574844074844075, "percentage": 3.29, "elapsed_time": "0:02:56", "remaining_time": "1:26:38", "throughput": 1362.9, "total_tokens": 240832}
+{"current_steps": 1270, "total_steps": 38480, "loss": 0.2256, "lr": 1.6489085239085238e-05, "epoch": 0.66008316008316, "percentage": 3.3, "elapsed_time": "0:02:57", "remaining_time": "1:26:37", "throughput": 1363.13, "total_tokens": 241824}
+{"current_steps": 1275, "total_steps": 38480, "loss": 0.485, "lr": 1.6554054054054053e-05, "epoch": 0.6626819126819127, "percentage": 3.31, "elapsed_time": "0:02:58", "remaining_time": "1:26:37", "throughput": 1363.36, "total_tokens": 242816}
+{"current_steps": 1280, "total_steps": 38480, "loss": 0.3629, "lr": 1.6619022869022873e-05, "epoch": 0.6652806652806653, "percentage": 3.33, "elapsed_time": "0:02:58", "remaining_time": "1:26:36", "throughput": 1363.59, "total_tokens": 243808}
+{"current_steps": 1285, "total_steps": 38480, "loss": 0.3958, "lr": 1.6683991683991685e-05, "epoch": 0.6678794178794178, "percentage": 3.34, "elapsed_time": "0:02:59", "remaining_time": "1:26:35", "throughput": 1363.47, "total_tokens": 244736}
+{"current_steps": 1290, "total_steps": 38480, "loss": 0.0782, "lr": 1.67489604989605e-05, "epoch": 0.6704781704781705, "percentage": 3.35, "elapsed_time": "0:03:00", "remaining_time": "1:26:35", "throughput": 1363.47, "total_tokens": 245696}
+{"current_steps": 1295, "total_steps": 38480, "loss": 0.4312, "lr": 1.6813929313929314e-05, "epoch": 0.6730769230769231, "percentage": 3.37, "elapsed_time": "0:03:00", "remaining_time": "1:26:34", "throughput": 1363.88, "total_tokens": 246720}
+{"current_steps": 1300, "total_steps": 38480, "loss": 0.2357, "lr": 1.687889812889813e-05, "epoch": 0.6756756756756757, "percentage": 3.38, "elapsed_time": "0:03:01", "remaining_time": "1:26:33", "throughput": 1364.31, "total_tokens": 247744}
+{"current_steps": 1305, "total_steps": 38480, "loss": 0.395, "lr": 1.6943866943866946e-05, "epoch": 0.6782744282744283, "percentage": 3.39, "elapsed_time": "0:03:02", "remaining_time": "1:26:32", "throughput": 1364.56, "total_tokens": 248736}
+{"current_steps": 1310, "total_steps": 38480, "loss": 0.3376, "lr": 1.7008835758835758e-05, "epoch": 0.6808731808731808, "percentage": 3.4, "elapsed_time": "0:03:02", "remaining_time": "1:26:31", "throughput": 1364.8, "total_tokens": 249728}
+{"current_steps": 1315, "total_steps": 38480, "loss": 0.4266, "lr": 1.7073804573804574e-05, "epoch": 0.6834719334719335, "percentage": 3.42, "elapsed_time": "0:03:03", "remaining_time": "1:26:30", "throughput": 1364.53, "total_tokens": 250624}
+{"current_steps": 1320, "total_steps": 38480, "loss": 0.1745, "lr": 1.713877338877339e-05, "epoch": 0.6860706860706861, "percentage": 3.43, "elapsed_time": "0:03:04", "remaining_time": "1:26:30", "throughput": 1364.74, "total_tokens": 251616}
+{"current_steps": 1325, "total_steps": 38480, "loss": 0.1987, "lr": 1.7203742203742202e-05, "epoch": 0.6886694386694386, "percentage": 3.44, "elapsed_time": "0:03:05", "remaining_time": "1:26:29", "throughput": 1364.26, "total_tokens": 252480}
+{"current_steps": 1330, "total_steps": 38480, "loss": 0.2202, "lr": 1.7268711018711022e-05, "epoch": 0.6912681912681913, "percentage": 3.46, "elapsed_time": "0:03:05", "remaining_time": "1:26:28", "throughput": 1364.3, "total_tokens": 253440}
+{"current_steps": 1335, "total_steps": 38480, "loss": 0.5486, "lr": 1.7333679833679834e-05, "epoch": 0.6938669438669439, "percentage": 3.47, "elapsed_time": "0:03:06", "remaining_time": "1:26:28", "throughput": 1364.34, "total_tokens": 254400}
+{"current_steps": 1340, "total_steps": 38480, "loss": 0.1507, "lr": 1.739864864864865e-05, "epoch": 0.6964656964656964, "percentage": 3.48, "elapsed_time": "0:03:07", "remaining_time": "1:26:27", "throughput": 1364.73, "total_tokens": 255424}
+{"current_steps": 1345, "total_steps": 38480, "loss": 0.1084, "lr": 1.7463617463617466e-05, "epoch": 0.6990644490644491, "percentage": 3.5, "elapsed_time": "0:03:07", "remaining_time": "1:26:26", "throughput": 1364.44, "total_tokens": 256320}
+{"current_steps": 1350, "total_steps": 38480, "loss": 0.3139, "lr": 1.752858627858628e-05, "epoch": 0.7016632016632016, "percentage": 3.51, "elapsed_time": "0:03:08", "remaining_time": "1:26:25", "throughput": 1364.14, "total_tokens": 257216}
+{"current_steps": 1355, "total_steps": 38480, "loss": 0.6216, "lr": 1.7593555093555095e-05, "epoch": 0.7042619542619543, "percentage": 3.52, "elapsed_time": "0:03:09", "remaining_time": "1:26:25", "throughput": 1363.69, "total_tokens": 258080}
+{"current_steps": 1360, "total_steps": 38480, "loss": 0.3345, "lr": 1.765852390852391e-05, "epoch": 0.7068607068607069, "percentage": 3.53, "elapsed_time": "0:03:09", "remaining_time": "1:26:24", "throughput": 1363.57, "total_tokens": 259008}
+{"current_steps": 1365, "total_steps": 38480, "loss": 0.1141, "lr": 1.7723492723492723e-05, "epoch": 0.7094594594594594, "percentage": 3.55, "elapsed_time": "0:03:10", "remaining_time": "1:26:23", "throughput": 1363.62, "total_tokens": 259968}
+{"current_steps": 1370, "total_steps": 38480, "loss": 0.1835, "lr": 1.778846153846154e-05, "epoch": 0.7120582120582121, "percentage": 3.56, "elapsed_time": "0:03:11", "remaining_time": "1:26:22", "throughput": 1363.86, "total_tokens": 260960}
+{"current_steps": 1375, "total_steps": 38480, "loss": 0.5223, "lr": 1.785343035343035e-05, "epoch": 0.7146569646569647, "percentage": 3.57, "elapsed_time": "0:03:12", "remaining_time": "1:26:22", "throughput": 1363.9, "total_tokens": 261920}
+{"current_steps": 1380, "total_steps": 38480, "loss": 0.1814, "lr": 1.791839916839917e-05, "epoch": 0.7172557172557172, "percentage": 3.59, "elapsed_time": "0:03:12", "remaining_time": "1:26:21", "throughput": 1363.61, "total_tokens": 262816}
+{"current_steps": 1385, "total_steps": 38480, "loss": 0.5034, "lr": 1.7983367983367987e-05, "epoch": 0.7198544698544699, "percentage": 3.6, "elapsed_time": "0:03:13", "remaining_time": "1:26:20", "throughput": 1363.86, "total_tokens": 263808}
+{"current_steps": 1390, "total_steps": 38480, "loss": 0.1949, "lr": 1.80483367983368e-05, "epoch": 0.7224532224532224, "percentage": 3.61, "elapsed_time": "0:03:14", "remaining_time": "1:26:19", "throughput": 1363.77, "total_tokens": 264736}
+{"current_steps": 1395, "total_steps": 38480, "loss": 0.2221, "lr": 1.8113305613305615e-05, "epoch": 0.725051975051975, "percentage": 3.63, "elapsed_time": "0:03:14", "remaining_time": "1:26:19", "throughput": 1363.84, "total_tokens": 265696}
+{"current_steps": 1400, "total_steps": 38480, "loss": 0.4934, "lr": 1.817827442827443e-05, "epoch": 0.7276507276507277, "percentage": 3.64, "elapsed_time": "0:03:15", "remaining_time": "1:26:18", "throughput": 1363.7, "total_tokens": 266624}
+{"current_steps": 1405, "total_steps": 38480, "loss": 0.4552, "lr": 1.8243243243243244e-05, "epoch": 0.7302494802494802, "percentage": 3.65, "elapsed_time": "0:03:16", "remaining_time": "1:26:17", "throughput": 1363.78, "total_tokens": 267584}
+{"current_steps": 1410, "total_steps": 38480, "loss": 0.2776, "lr": 1.830821205821206e-05, "epoch": 0.7328482328482329, "percentage": 3.66, "elapsed_time": "0:03:16", "remaining_time": "1:26:16", "throughput": 1363.85, "total_tokens": 268544}
+{"current_steps": 1415, "total_steps": 38480, "loss": 0.1937, "lr": 1.8373180873180872e-05, "epoch": 0.7354469854469855, "percentage": 3.68, "elapsed_time": "0:03:17", "remaining_time": "1:26:15", "throughput": 1363.77, "total_tokens": 269472}
+{"current_steps": 1420, "total_steps": 38480, "loss": 0.222, "lr": 1.8438149688149688e-05, "epoch": 0.738045738045738, "percentage": 3.69, "elapsed_time": "0:03:18", "remaining_time": "1:26:15", "throughput": 1363.52, "total_tokens": 270368}
+{"current_steps": 1425, "total_steps": 38480, "loss": 0.3261, "lr": 1.8503118503118504e-05, "epoch": 0.7406444906444907, "percentage": 3.7, "elapsed_time": "0:03:18", "remaining_time": "1:26:14", "throughput": 1363.91, "total_tokens": 271392}
+{"current_steps": 1430, "total_steps": 38480, "loss": 0.1954, "lr": 1.856808731808732e-05, "epoch": 0.7432432432432432, "percentage": 3.72, "elapsed_time": "0:03:19", "remaining_time": "1:26:13", "throughput": 1364.14, "total_tokens": 272384}
+{"current_steps": 1435, "total_steps": 38480, "loss": 0.1465, "lr": 1.8633056133056136e-05, "epoch": 0.7458419958419958, "percentage": 3.73, "elapsed_time": "0:03:20", "remaining_time": "1:26:12", "throughput": 1363.9, "total_tokens": 273280}
+{"current_steps": 1440, "total_steps": 38480, "loss": 0.0706, "lr": 1.8698024948024952e-05, "epoch": 0.7484407484407485, "percentage": 3.74, "elapsed_time": "0:03:21", "remaining_time": "1:26:11", "throughput": 1363.66, "total_tokens": 274176}
+{"current_steps": 1445, "total_steps": 38480, "loss": 0.1589, "lr": 1.8762993762993764e-05, "epoch": 0.751039501039501, "percentage": 3.76, "elapsed_time": "0:03:21", "remaining_time": "1:26:11", "throughput": 1363.68, "total_tokens": 275136}
+{"current_steps": 1450, "total_steps": 38480, "loss": 0.3379, "lr": 1.882796257796258e-05, "epoch": 0.7536382536382537, "percentage": 3.77, "elapsed_time": "0:03:22", "remaining_time": "1:26:10", "throughput": 1363.88, "total_tokens": 276128}
+{"current_steps": 1455, "total_steps": 38480, "loss": 0.2128, "lr": 1.8892931392931393e-05, "epoch": 0.7562370062370062, "percentage": 3.78, "elapsed_time": "0:03:23", "remaining_time": "1:26:09", "throughput": 1363.61, "total_tokens": 277024}
+{"current_steps": 1460, "total_steps": 38480, "loss": 0.253, "lr": 1.895790020790021e-05, "epoch": 0.7588357588357588, "percentage": 3.79, "elapsed_time": "0:03:23", "remaining_time": "1:26:08", "throughput": 1363.96, "total_tokens": 278048}
+{"current_steps": 1465, "total_steps": 38480, "loss": 0.1789, "lr": 1.9022869022869025e-05, "epoch": 0.7614345114345115, "percentage": 3.81, "elapsed_time": "0:03:24", "remaining_time": "1:26:08", "throughput": 1364.16, "total_tokens": 279040}
+{"current_steps": 1470, "total_steps": 38480, "loss": 0.2861, "lr": 1.9087837837837837e-05, "epoch": 0.764033264033264, "percentage": 3.82, "elapsed_time": "0:03:25", "remaining_time": "1:26:07", "throughput": 1364.19, "total_tokens": 280000}
+{"current_steps": 1475, "total_steps": 38480, "loss": 0.0348, "lr": 1.9152806652806653e-05, "epoch": 0.7666320166320166, "percentage": 3.83, "elapsed_time": "0:03:25", "remaining_time": "1:26:06", "throughput": 1364.23, "total_tokens": 280960}
+{"current_steps": 1480, "total_steps": 38480, "loss": 0.2644, "lr": 1.921777546777547e-05, "epoch": 0.7692307692307693, "percentage": 3.85, "elapsed_time": "0:03:26", "remaining_time": "1:26:06", "throughput": 1364.3, "total_tokens": 281920}
+{"current_steps": 1485, "total_steps": 38480, "loss": 0.176, "lr": 1.9282744282744285e-05, "epoch": 0.7718295218295218, "percentage": 3.86, "elapsed_time": "0:03:27", "remaining_time": "1:26:05", "throughput": 1364.67, "total_tokens": 282944}
+{"current_steps": 1490, "total_steps": 38480, "loss": 0.2196, "lr": 1.93477130977131e-05, "epoch": 0.7744282744282744, "percentage": 3.87, "elapsed_time": "0:03:28", "remaining_time": "1:26:04", "throughput": 1364.59, "total_tokens": 283872}
+{"current_steps": 1495, "total_steps": 38480, "loss": 0.3736, "lr": 1.9412681912681913e-05, "epoch": 0.777027027027027, "percentage": 3.89, "elapsed_time": "0:03:28", "remaining_time": "1:26:03", "throughput": 1364.52, "total_tokens": 284800}
+{"current_steps": 1500, "total_steps": 38480, "loss": 0.4256, "lr": 1.947765072765073e-05, "epoch": 0.7796257796257796, "percentage": 3.9, "elapsed_time": "0:03:29", "remaining_time": "1:26:02", "throughput": 1364.13, "total_tokens": 285664}
+{"current_steps": 1505, "total_steps": 38480, "loss": 0.0563, "lr": 1.9542619542619545e-05, "epoch": 0.7822245322245323, "percentage": 3.91, "elapsed_time": "0:03:30", "remaining_time": "1:26:02", "throughput": 1364.15, "total_tokens": 286624}
+{"current_steps": 1510, "total_steps": 38480, "loss": 0.1873, "lr": 1.9607588357588358e-05, "epoch": 0.7848232848232848, "percentage": 3.92, "elapsed_time": "0:03:30", "remaining_time": "1:26:01", "throughput": 1363.89, "total_tokens": 287520}
+{"current_steps": 1515, "total_steps": 38480, "loss": 0.1794, "lr": 1.9672557172557174e-05, "epoch": 0.7874220374220374, "percentage": 3.94, "elapsed_time": "0:03:31", "remaining_time": "1:26:00", "throughput": 1364.12, "total_tokens": 288512}
+{"current_steps": 1520, "total_steps": 38480, "loss": 0.3425, "lr": 1.9737525987525986e-05, "epoch": 0.7900207900207901, "percentage": 3.95, "elapsed_time": "0:03:32", "remaining_time": "1:25:59", "throughput": 1364.19, "total_tokens": 289472}
+{"current_steps": 1525, "total_steps": 38480, "loss": 0.1134, "lr": 1.9802494802494802e-05, "epoch": 0.7926195426195426, "percentage": 3.96, "elapsed_time": "0:03:32", "remaining_time": "1:25:58", "throughput": 1364.1, "total_tokens": 290400}
+{"current_steps": 1530, "total_steps": 38480, "loss": 0.1463, "lr": 1.9867463617463618e-05, "epoch": 0.7952182952182952, "percentage": 3.98, "elapsed_time": "0:03:33", "remaining_time": "1:25:58", "throughput": 1364.14, "total_tokens": 291360}
+{"current_steps": 1535, "total_steps": 38480, "loss": 0.2242, "lr": 1.9932432432432434e-05, "epoch": 0.7978170478170478, "percentage": 3.99, "elapsed_time": "0:03:34", "remaining_time": "1:25:57", "throughput": 1364.51, "total_tokens": 292384}
+{"current_steps": 1540, "total_steps": 38480, "loss": 0.3079, "lr": 1.999740124740125e-05, "epoch": 0.8004158004158004, "percentage": 4.0, "elapsed_time": "0:03:34", "remaining_time": "1:25:56", "throughput": 1365.17, "total_tokens": 293472}
+{"current_steps": 1545, "total_steps": 38480, "loss": 0.2144, "lr": 2.0062370062370066e-05, "epoch": 0.803014553014553, "percentage": 4.02, "elapsed_time": "0:03:35", "remaining_time": "1:25:55", "throughput": 1365.22, "total_tokens": 294432}
+{"current_steps": 1550, "total_steps": 38480, "loss": 0.4301, "lr": 2.0127338877338878e-05, "epoch": 0.8056133056133056, "percentage": 4.03, "elapsed_time": "0:03:36", "remaining_time": "1:25:54", "throughput": 1365.58, "total_tokens": 295456}
+{"current_steps": 1555, "total_steps": 38480, "loss": 0.1254, "lr": 2.0192307692307694e-05, "epoch": 0.8082120582120582, "percentage": 4.04, "elapsed_time": "0:03:37", "remaining_time": "1:25:54", "throughput": 1365.79, "total_tokens": 296448}
+{"current_steps": 1560, "total_steps": 38480, "loss": 0.073, "lr": 2.0257276507276507e-05, "epoch": 0.8108108108108109, "percentage": 4.05, "elapsed_time": "0:03:37", "remaining_time": "1:25:53", "throughput": 1365.71, "total_tokens": 297376}
+{"current_steps": 1565, "total_steps": 38480, "loss": 0.1712, "lr": 2.0322245322245323e-05, "epoch": 0.8134095634095634, "percentage": 4.07, "elapsed_time": "0:03:38", "remaining_time": "1:25:52", "throughput": 1365.77, "total_tokens": 298336}
+{"current_steps": 1570, "total_steps": 38480, "loss": 0.0883, "lr": 2.038721413721414e-05, "epoch": 0.816008316008316, "percentage": 4.08, "elapsed_time": "0:03:39", "remaining_time": "1:25:51", "throughput": 1365.69, "total_tokens": 299264}
+{"current_steps": 1575, "total_steps": 38480, "loss": 0.1459, "lr": 2.045218295218295e-05, "epoch": 0.8186070686070686, "percentage": 4.09, "elapsed_time": "0:03:39", "remaining_time": "1:25:50", "throughput": 1365.9, "total_tokens": 300256}
+{"current_steps": 1580, "total_steps": 38480, "loss": 0.0455, "lr": 2.0517151767151767e-05, "epoch": 0.8212058212058212, "percentage": 4.11, "elapsed_time": "0:03:40", "remaining_time": "1:25:50", "throughput": 1365.79, "total_tokens": 301184}
+{"current_steps": 1585, "total_steps": 38480, "loss": 0.1223, "lr": 2.0582120582120583e-05, "epoch": 0.8238045738045738, "percentage": 4.12, "elapsed_time": "0:03:41", "remaining_time": "1:25:49", "throughput": 1365.85, "total_tokens": 302144}
+{"current_steps": 1590, "total_steps": 38480, "loss": 0.2562, "lr": 2.06470893970894e-05, "epoch": 0.8264033264033264, "percentage": 4.13, "elapsed_time": "0:03:41", "remaining_time": "1:25:48", "throughput": 1366.03, "total_tokens": 303136}
+{"current_steps": 1595, "total_steps": 38480, "loss": 0.2903, "lr": 2.0712058212058215e-05, "epoch": 0.829002079002079, "percentage": 4.15, "elapsed_time": "0:03:42", "remaining_time": "1:25:47", "throughput": 1366.08, "total_tokens": 304096}
+{"current_steps": 1600, "total_steps": 38480, "loss": 0.1175, "lr": 2.0777027027027027e-05, "epoch": 0.8316008316008316, "percentage": 4.16, "elapsed_time": "0:03:43", "remaining_time": "1:25:46", "throughput": 1365.86, "total_tokens": 304992}
+{"current_steps": 1605, "total_steps": 38480, "loss": 0.441, "lr": 2.0841995841995843e-05, "epoch": 0.8341995841995842, "percentage": 4.17, "elapsed_time": "0:03:43", "remaining_time": "1:25:46", "throughput": 1366.06, "total_tokens": 305984}
+{"current_steps": 1610, "total_steps": 38480, "loss": 0.2397, "lr": 2.090696465696466e-05, "epoch": 0.8367983367983368, "percentage": 4.18, "elapsed_time": "0:03:44", "remaining_time": "1:25:45", "throughput": 1366.39, "total_tokens": 307008}
+{"current_steps": 1615, "total_steps": 38480, "loss": 0.1155, "lr": 2.097193347193347e-05, "epoch": 0.8393970893970893, "percentage": 4.2, "elapsed_time": "0:03:45", "remaining_time": "1:25:44", "throughput": 1366.26, "total_tokens": 307936}
+{"current_steps": 1620, "total_steps": 38480, "loss": 0.1308, "lr": 2.1036902286902288e-05, "epoch": 0.841995841995842, "percentage": 4.21, "elapsed_time": "0:03:46", "remaining_time": "1:25:44", "throughput": 1366.31, "total_tokens": 308896}
+{"current_steps": 1625, "total_steps": 38480, "loss": 0.3235, "lr": 2.1101871101871104e-05, "epoch": 0.8445945945945946, "percentage": 4.22, "elapsed_time": "0:03:46", "remaining_time": "1:25:43", "throughput": 1366.22, "total_tokens": 309824}
+{"current_steps": 1630, "total_steps": 38480, "loss": 0.2889, "lr": 2.1166839916839916e-05, "epoch": 0.8471933471933472, "percentage": 4.24, "elapsed_time": "0:03:47", "remaining_time": "1:25:42", "throughput": 1366.68, "total_tokens": 310880}
+{"current_steps": 1635, "total_steps": 38480, "loss": 0.37, "lr": 2.1231808731808732e-05, "epoch": 0.8497920997920998, "percentage": 4.25, "elapsed_time": "0:03:48", "remaining_time": "1:25:41", "throughput": 1366.73, "total_tokens": 311840}
+{"current_steps": 1640, "total_steps": 38480, "loss": 0.093, "lr": 2.1296777546777548e-05, "epoch": 0.8523908523908524, "percentage": 4.26, "elapsed_time": "0:03:48", "remaining_time": "1:25:40", "throughput": 1366.77, "total_tokens": 312800}
+{"current_steps": 1645, "total_steps": 38480, "loss": 0.2817, "lr": 2.1361746361746364e-05, "epoch": 0.854989604989605, "percentage": 4.27, "elapsed_time": "0:03:49", "remaining_time": "1:25:40", "throughput": 1366.96, "total_tokens": 313792}
+{"current_steps": 1650, "total_steps": 38480, "loss": 0.1116, "lr": 2.142671517671518e-05, "epoch": 0.8575883575883576, "percentage": 4.29, "elapsed_time": "0:03:50", "remaining_time": "1:25:39", "throughput": 1366.87, "total_tokens": 314720}
+{"current_steps": 1655, "total_steps": 38480, "loss": 0.1472, "lr": 2.1491683991683992e-05, "epoch": 0.8601871101871101, "percentage": 4.3, "elapsed_time": "0:03:50", "remaining_time": "1:25:38", "throughput": 1366.79, "total_tokens": 315648}
+{"current_steps": 1660, "total_steps": 38480, "loss": 0.0935, "lr": 2.1556652806652808e-05, "epoch": 0.8627858627858628, "percentage": 4.31, "elapsed_time": "0:03:51", "remaining_time": "1:25:37", "throughput": 1366.85, "total_tokens": 316608}
+{"current_steps": 1665, "total_steps": 38480, "loss": 0.3108, "lr": 2.1621621621621624e-05, "epoch": 0.8653846153846154, "percentage": 4.33, "elapsed_time": "0:03:52", "remaining_time": "1:25:36", "throughput": 1367.04, "total_tokens": 317600}
+{"current_steps": 1670, "total_steps": 38480, "loss": 0.5784, "lr": 2.1686590436590437e-05, "epoch": 0.867983367983368, "percentage": 4.34, "elapsed_time": "0:03:53", "remaining_time": "1:25:36", "throughput": 1367.09, "total_tokens": 318560}
+{"current_steps": 1675, "total_steps": 38480, "loss": 0.1561, "lr": 2.1751559251559253e-05, "epoch": 0.8705821205821206, "percentage": 4.35, "elapsed_time": "0:03:53", "remaining_time": "1:25:35", "throughput": 1366.87, "total_tokens": 319456}
+{"current_steps": 1680, "total_steps": 38480, "loss": 0.2432, "lr": 2.1816528066528065e-05, "epoch": 0.8731808731808732, "percentage": 4.37, "elapsed_time": "0:03:54", "remaining_time": "1:25:34", "throughput": 1366.78, "total_tokens": 320384}
+{"current_steps": 1685, "total_steps": 38480, "loss": 0.2376, "lr": 2.188149688149688e-05, "epoch": 0.8757796257796258, "percentage": 4.38, "elapsed_time": "0:03:55", "remaining_time": "1:25:33", "throughput": 1366.8, "total_tokens": 321344}
+{"current_steps": 1690, "total_steps": 38480, "loss": 0.1183, "lr": 2.1946465696465697e-05, "epoch": 0.8783783783783784, "percentage": 4.39, "elapsed_time": "0:03:55", "remaining_time": "1:25:33", "throughput": 1366.97, "total_tokens": 322336}
+{"current_steps": 1695, "total_steps": 38480, "loss": 0.1023, "lr": 2.2011434511434513e-05, "epoch": 0.8809771309771309, "percentage": 4.4, "elapsed_time": "0:03:56", "remaining_time": "1:25:32", "throughput": 1367.27, "total_tokens": 323360}
+{"current_steps": 1700, "total_steps": 38480, "loss": 0.1696, "lr": 2.207640332640333e-05, "epoch": 0.8835758835758836, "percentage": 4.42, "elapsed_time": "0:03:57", "remaining_time": "1:25:31", "throughput": 1367.05, "total_tokens": 324256}
+{"current_steps": 1705, "total_steps": 38480, "loss": 0.3673, "lr": 2.2141372141372145e-05, "epoch": 0.8861746361746362, "percentage": 4.43, "elapsed_time": "0:03:57", "remaining_time": "1:25:30", "throughput": 1366.84, "total_tokens": 325152}
+{"current_steps": 1710, "total_steps": 38480, "loss": 0.2393, "lr": 2.2206340956340957e-05, "epoch": 0.8887733887733887, "percentage": 4.44, "elapsed_time": "0:03:58", "remaining_time": "1:25:30", "throughput": 1366.63, "total_tokens": 326048}
+{"current_steps": 1715, "total_steps": 38480, "loss": 0.1194, "lr": 2.2271309771309773e-05, "epoch": 0.8913721413721414, "percentage": 4.46, "elapsed_time": "0:03:59", "remaining_time": "1:25:29", "throughput": 1366.68, "total_tokens": 327008}
+{"current_steps": 1720, "total_steps": 38480, "loss": 0.083, "lr": 2.2336278586278586e-05, "epoch": 0.893970893970894, "percentage": 4.47, "elapsed_time": "0:03:59", "remaining_time": "1:25:28", "throughput": 1366.59, "total_tokens": 327936}
+{"current_steps": 1725, "total_steps": 38480, "loss": 0.3519, "lr": 2.24012474012474e-05, "epoch": 0.8965696465696466, "percentage": 4.48, "elapsed_time": "0:04:00", "remaining_time": "1:25:27", "throughput": 1366.59, "total_tokens": 328896}
+{"current_steps": 1730, "total_steps": 38480, "loss": 0.1444, "lr": 2.2466216216216218e-05, "epoch": 0.8991683991683992, "percentage": 4.5, "elapsed_time": "0:04:01", "remaining_time": "1:25:27", "throughput": 1366.76, "total_tokens": 329888}
+{"current_steps": 1735, "total_steps": 38480, "loss": 0.1719, "lr": 2.253118503118503e-05, "epoch": 0.9017671517671517, "percentage": 4.51, "elapsed_time": "0:04:02", "remaining_time": "1:25:26", "throughput": 1366.78, "total_tokens": 330848}
+{"current_steps": 1740, "total_steps": 38480, "loss": 0.2148, "lr": 2.2596153846153846e-05, "epoch": 0.9043659043659044, "percentage": 4.52, "elapsed_time": "0:04:02", "remaining_time": "1:25:25", "throughput": 1366.54, "total_tokens": 331744}
+{"current_steps": 1745, "total_steps": 38480, "loss": 0.2771, "lr": 2.2661122661122662e-05, "epoch": 0.906964656964657, "percentage": 4.53, "elapsed_time": "0:04:03", "remaining_time": "1:25:25", "throughput": 1366.7, "total_tokens": 332736}
+{"current_steps": 1750, "total_steps": 38480, "loss": 0.1562, "lr": 2.2726091476091478e-05, "epoch": 0.9095634095634095, "percentage": 4.55, "elapsed_time": "0:04:04", "remaining_time": "1:25:24", "throughput": 1366.86, "total_tokens": 333728}
+{"current_steps": 1755, "total_steps": 38480, "loss": 0.0393, "lr": 2.2791060291060294e-05, "epoch": 0.9121621621621622, "percentage": 4.56, "elapsed_time": "0:04:04", "remaining_time": "1:25:23", "throughput": 1366.63, "total_tokens": 334624}
+{"current_steps": 1760, "total_steps": 38480, "loss": 0.1265, "lr": 2.2856029106029106e-05, "epoch": 0.9147609147609148, "percentage": 4.57, "elapsed_time": "0:04:05", "remaining_time": "1:25:22", "throughput": 1366.81, "total_tokens": 335616}
+{"current_steps": 1765, "total_steps": 38480, "loss": 0.2743, "lr": 2.2920997920997922e-05, "epoch": 0.9173596673596673, "percentage": 4.59, "elapsed_time": "0:04:06", "remaining_time": "1:25:22", "throughput": 1366.73, "total_tokens": 336544}
+{"current_steps": 1770, "total_steps": 38480, "loss": 0.2137, "lr": 2.2985966735966738e-05, "epoch": 0.91995841995842, "percentage": 4.6, "elapsed_time": "0:04:06", "remaining_time": "1:25:21", "throughput": 1366.65, "total_tokens": 337472}
+{"current_steps": 1775, "total_steps": 38480, "loss": 0.3174, "lr": 2.305093555093555e-05, "epoch": 0.9225571725571725, "percentage": 4.61, "elapsed_time": "0:04:07", "remaining_time": "1:25:20", "throughput": 1366.55, "total_tokens": 338400}
+{"current_steps": 1780, "total_steps": 38480, "loss": 0.1148, "lr": 2.3115904365904367e-05, "epoch": 0.9251559251559252, "percentage": 4.63, "elapsed_time": "0:04:08", "remaining_time": "1:25:20", "throughput": 1366.58, "total_tokens": 339360}
+{"current_steps": 1785, "total_steps": 38480, "loss": 0.2128, "lr": 2.3180873180873183e-05, "epoch": 0.9277546777546778, "percentage": 4.64, "elapsed_time": "0:04:09", "remaining_time": "1:25:19", "throughput": 1366.47, "total_tokens": 340288}
+{"current_steps": 1790, "total_steps": 38480, "loss": 0.2603, "lr": 2.3245841995841995e-05, "epoch": 0.9303534303534303, "percentage": 4.65, "elapsed_time": "0:04:09", "remaining_time": "1:25:18", "throughput": 1366.74, "total_tokens": 341312}
+{"current_steps": 1795, "total_steps": 38480, "loss": 0.351, "lr": 2.331081081081081e-05, "epoch": 0.932952182952183, "percentage": 4.66, "elapsed_time": "0:04:10", "remaining_time": "1:25:18", "throughput": 1366.76, "total_tokens": 342272}
+{"current_steps": 1800, "total_steps": 38480, "loss": 0.3326, "lr": 2.3375779625779627e-05, "epoch": 0.9355509355509356, "percentage": 4.68, "elapsed_time": "0:04:11", "remaining_time": "1:25:17", "throughput": 1366.41, "total_tokens": 343136}
+{"current_steps": 1805, "total_steps": 38480, "loss": 0.2573, "lr": 2.3440748440748443e-05, "epoch": 0.9381496881496881, "percentage": 4.69, "elapsed_time": "0:04:11", "remaining_time": "1:25:16", "throughput": 1366.31, "total_tokens": 344064}
+{"current_steps": 1810, "total_steps": 38480, "loss": 0.1404, "lr": 2.350571725571726e-05, "epoch": 0.9407484407484408, "percentage": 4.7, "elapsed_time": "0:04:12", "remaining_time": "1:25:15", "throughput": 1366.33, "total_tokens": 345024}
+{"current_steps": 1815, "total_steps": 38480, "loss": 0.2378, "lr": 2.357068607068607e-05, "epoch": 0.9433471933471933, "percentage": 4.72, "elapsed_time": "0:04:13", "remaining_time": "1:25:15", "throughput": 1366.48, "total_tokens": 346016}
+{"current_steps": 1820, "total_steps": 38480, "loss": 0.3027, "lr": 2.3635654885654887e-05, "epoch": 0.9459459459459459, "percentage": 4.73, "elapsed_time": "0:04:13", "remaining_time": "1:25:14", "throughput": 1366.62, "total_tokens": 347008}
+{"current_steps": 1825, "total_steps": 38480, "loss": 0.1569, "lr": 2.3700623700623703e-05, "epoch": 0.9485446985446986, "percentage": 4.74, "elapsed_time": "0:04:14", "remaining_time": "1:25:13", "throughput": 1366.54, "total_tokens": 347936}
+{"current_steps": 1830, "total_steps": 38480, "loss": 0.185, "lr": 2.3765592515592516e-05, "epoch": 0.9511434511434511, "percentage": 4.76, "elapsed_time": "0:04:15", "remaining_time": "1:25:13", "throughput": 1366.54, "total_tokens": 348896}
+{"current_steps": 1835, "total_steps": 38480, "loss": 0.2441, "lr": 2.383056133056133e-05, "epoch": 0.9537422037422038, "percentage": 4.77, "elapsed_time": "0:04:16", "remaining_time": "1:25:12", "throughput": 1366.45, "total_tokens": 349824}
+{"current_steps": 1840, "total_steps": 38480, "loss": 0.1772, "lr": 2.3895530145530144e-05, "epoch": 0.9563409563409564, "percentage": 4.78, "elapsed_time": "0:04:16", "remaining_time": "1:25:11", "throughput": 1366.34, "total_tokens": 350752}
+{"current_steps": 1845, "total_steps": 38480, "loss": 0.1803, "lr": 2.396049896049896e-05, "epoch": 0.9589397089397089, "percentage": 4.79, "elapsed_time": "0:04:17", "remaining_time": "1:25:11", "throughput": 1366.12, "total_tokens": 351648}
+{"current_steps": 1850, "total_steps": 38480, "loss": 0.1999, "lr": 2.4025467775467776e-05, "epoch": 0.9615384615384616, "percentage": 4.81, "elapsed_time": "0:04:18", "remaining_time": "1:25:10", "throughput": 1366.16, "total_tokens": 352608}
+{"current_steps": 1855, "total_steps": 38480, "loss": 0.2177, "lr": 2.4090436590436592e-05, "epoch": 0.9641372141372141, "percentage": 4.82, "elapsed_time": "0:04:18", "remaining_time": "1:25:09", "throughput": 1366.44, "total_tokens": 353632}
+{"current_steps": 1860, "total_steps": 38480, "loss": 0.0825, "lr": 2.4155405405405408e-05, "epoch": 0.9667359667359667, "percentage": 4.83, "elapsed_time": "0:04:19", "remaining_time": "1:25:08", "throughput": 1366.47, "total_tokens": 354592}
+{"current_steps": 1865, "total_steps": 38480, "loss": 0.288, "lr": 2.4220374220374224e-05, "epoch": 0.9693347193347194, "percentage": 4.85, "elapsed_time": "0:04:20", "remaining_time": "1:25:08", "throughput": 1366.12, "total_tokens": 355456}
+{"current_steps": 1870, "total_steps": 38480, "loss": 0.1211, "lr": 2.4285343035343036e-05, "epoch": 0.9719334719334719, "percentage": 4.86, "elapsed_time": "0:04:20", "remaining_time": "1:25:07", "throughput": 1366.29, "total_tokens": 356448}
+{"current_steps": 1875, "total_steps": 38480, "loss": 0.3246, "lr": 2.4350311850311852e-05, "epoch": 0.9745322245322245, "percentage": 4.87, "elapsed_time": "0:04:21", "remaining_time": "1:25:06", "throughput": 1366.34, "total_tokens": 357408}
+{"current_steps": 1880, "total_steps": 38480, "loss": 0.1601, "lr": 2.4415280665280665e-05, "epoch": 0.9771309771309772, "percentage": 4.89, "elapsed_time": "0:04:22", "remaining_time": "1:25:05", "throughput": 1366.39, "total_tokens": 358368}
+{"current_steps": 1885, "total_steps": 38480, "loss": 0.2995, "lr": 2.448024948024948e-05, "epoch": 0.9797297297297297, "percentage": 4.9, "elapsed_time": "0:04:22", "remaining_time": "1:25:05", "throughput": 1366.32, "total_tokens": 359296}
+{"current_steps": 1890, "total_steps": 38480, "loss": 0.1981, "lr": 2.4545218295218297e-05, "epoch": 0.9823284823284824, "percentage": 4.91, "elapsed_time": "0:04:23", "remaining_time": "1:25:04", "throughput": 1366.38, "total_tokens": 360256}
+{"current_steps": 1895, "total_steps": 38480, "loss": 0.107, "lr": 2.461018711018711e-05, "epoch": 0.9849272349272349, "percentage": 4.92, "elapsed_time": "0:04:24", "remaining_time": "1:25:03", "throughput": 1366.43, "total_tokens": 361216}
+{"current_steps": 1900, "total_steps": 38480, "loss": 0.2286, "lr": 2.4675155925155925e-05, "epoch": 0.9875259875259875, "percentage": 4.94, "elapsed_time": "0:04:25", "remaining_time": "1:25:02", "throughput": 1366.48, "total_tokens": 362176}
+{"current_steps": 1905, "total_steps": 38480, "loss": 0.3144, "lr": 2.4740124740124744e-05, "epoch": 0.9901247401247402, "percentage": 4.95, "elapsed_time": "0:04:25", "remaining_time": "1:25:01", "throughput": 1366.77, "total_tokens": 363200}
+{"current_steps": 1910, "total_steps": 38480, "loss": 0.0685, "lr": 2.4805093555093557e-05, "epoch": 0.9927234927234927, "percentage": 4.96, "elapsed_time": "0:04:26", "remaining_time": "1:25:01", "throughput": 1366.78, "total_tokens": 364160}
+{"current_steps": 1915, "total_steps": 38480, "loss": 0.2603, "lr": 2.4870062370062373e-05, "epoch": 0.9953222453222453, "percentage": 4.98, "elapsed_time": "0:04:27", "remaining_time": "1:25:00", "throughput": 1366.45, "total_tokens": 365024}
+{"current_steps": 1920, "total_steps": 38480, "loss": 0.343, "lr": 2.4935031185031185e-05, "epoch": 0.997920997920998, "percentage": 4.99, "elapsed_time": "0:04:27", "remaining_time": "1:24:59", "throughput": 1366.6, "total_tokens": 366016}
+{"current_steps": 1924, "total_steps": 38480, "eval_loss": 0.20758913457393646, "epoch": 1.0, "percentage": 5.0, "elapsed_time": "0:04:36", "remaining_time": "1:27:32", "throughput": 1326.55, "total_tokens": 366712}
+{"current_steps": 1925, "total_steps": 38480, "loss": 0.3316, "lr": 2.5e-05, "epoch": 1.0005197505197505, "percentage": 5.0, "elapsed_time": "0:04:37", "remaining_time": "1:27:52", "throughput": 1321.43, "total_tokens": 366904}
+{"current_steps": 1930, "total_steps": 38480, "loss": 0.2137, "lr": 2.5064968814968814e-05, "epoch": 1.003118503118503, "percentage": 5.02, "elapsed_time": "0:04:38", "remaining_time": "1:27:52", "throughput": 1321.24, "total_tokens": 367832}
+{"current_steps": 1935, "total_steps": 38480, "loss": 0.1894, "lr": 2.5129937629937633e-05, "epoch": 1.0057172557172558, "percentage": 5.03, "elapsed_time": "0:04:39", "remaining_time": "1:28:02", "throughput": 1318.52, "total_tokens": 368824}
+{"current_steps": 1940, "total_steps": 38480, "loss": 0.4569, "lr": 2.5194906444906446e-05, "epoch": 1.0083160083160083, "percentage": 5.04, "elapsed_time": "0:04:40", "remaining_time": "1:28:02", "throughput": 1318.72, "total_tokens": 369816}
+{"current_steps": 1945, "total_steps": 38480, "loss": 0.2239, "lr": 2.5259875259875258e-05, "epoch": 1.0109147609147608, "percentage": 5.05, "elapsed_time": "0:04:41", "remaining_time": "1:28:00", "throughput": 1318.83, "total_tokens": 370776}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:911770d6d8d9b313b4d4562fd1969362337bc8f34863aae0842be79db2b5b91d
+size 6289