Upload folder using huggingface_hub
Browse files
slots/1/checkpoint-6452/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"TwinyForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"hidden_dropout": 0.0,
|
| 8 |
+
"hidden_size": 768,
|
| 9 |
+
"initializer_range": 0.02,
|
| 10 |
+
"intermediate_size": 3072,
|
| 11 |
+
"max_position_embeddings": 128,
|
| 12 |
+
"model_type": "twiny",
|
| 13 |
+
"neftune_alpha": 0.0,
|
| 14 |
+
"num_attention_heads": 12,
|
| 15 |
+
"num_hidden_layers": 2,
|
| 16 |
+
"num_key_value_heads": 3,
|
| 17 |
+
"qk_norm": true,
|
| 18 |
+
"rezero_init": 1.0,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_theta": 10000.0,
|
| 21 |
+
"transformers_version": "5.0.0",
|
| 22 |
+
"use_cache": false,
|
| 23 |
+
"vocab_size": 32000
|
| 24 |
+
}
|
slots/1/checkpoint-6452/trainer_state.json
ADDED
|
@@ -0,0 +1,2295 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0257514941558634,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 6452,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0003140426116568692,
|
| 14 |
+
"grad_norm": 11.236513137817383,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 14.929291725158691,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0062808522331373835,
|
| 21 |
+
"grad_norm": 8.230694770812988,
|
| 22 |
+
"learning_rate": 4.9999363065476e-05,
|
| 23 |
+
"loss": 13.227896439401727,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.012561704466274767,
|
| 28 |
+
"grad_norm": 15.541277885437012,
|
| 29 |
+
"learning_rate": 4.999745229435885e-05,
|
| 30 |
+
"loss": 11.76095199584961,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.018842556699412153,
|
| 35 |
+
"grad_norm": 19.88605308532715,
|
| 36 |
+
"learning_rate": 4.999406489735261e-05,
|
| 37 |
+
"loss": 10.630082702636718,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.025123408932549534,
|
| 42 |
+
"grad_norm": 21.772388458251953,
|
| 43 |
+
"learning_rate": 4.998953978539926e-05,
|
| 44 |
+
"loss": 10.612107086181641,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03140426116568692,
|
| 49 |
+
"grad_norm": 24.13600730895996,
|
| 50 |
+
"learning_rate": 4.998374133739565e-05,
|
| 51 |
+
"loss": 8.763066864013672,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.037685113398824306,
|
| 56 |
+
"grad_norm": 33.56028747558594,
|
| 57 |
+
"learning_rate": 4.997666984880032e-05,
|
| 58 |
+
"loss": 11.96340103149414,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.04396596563196169,
|
| 63 |
+
"grad_norm": 51.29957962036133,
|
| 64 |
+
"learning_rate": 4.996785126662353e-05,
|
| 65 |
+
"loss": 7.662068939208984,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.05024681786509907,
|
| 70 |
+
"grad_norm": 34.75215530395508,
|
| 71 |
+
"learning_rate": 4.9957623010947197e-05,
|
| 72 |
+
"loss": 7.163829040527344,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.056527670098236456,
|
| 77 |
+
"grad_norm": 48.993743896484375,
|
| 78 |
+
"learning_rate": 4.994598565925399e-05,
|
| 79 |
+
"loss": 6.738508605957032,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.06280852233137384,
|
| 84 |
+
"grad_norm": 37.503150939941406,
|
| 85 |
+
"learning_rate": 4.9932939868583476e-05,
|
| 86 |
+
"loss": 5.927839660644532,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.06908937456451122,
|
| 91 |
+
"grad_norm": 34.306514739990234,
|
| 92 |
+
"learning_rate": 4.991848637549505e-05,
|
| 93 |
+
"loss": 5.548626327514649,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.07537022679764861,
|
| 98 |
+
"grad_norm": 25.989212036132812,
|
| 99 |
+
"learning_rate": 4.990345241504926e-05,
|
| 100 |
+
"loss": 5.030820083618164,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.08165107903078599,
|
| 105 |
+
"grad_norm": 23.606063842773438,
|
| 106 |
+
"learning_rate": 4.98862563217225e-05,
|
| 107 |
+
"loss": 4.923490524291992,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.08793193126392337,
|
| 112 |
+
"grad_norm": 36.173194885253906,
|
| 113 |
+
"learning_rate": 4.9867655161709595e-05,
|
| 114 |
+
"loss": 4.673915100097656,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.09421278349706076,
|
| 119 |
+
"grad_norm": 18.624378204345703,
|
| 120 |
+
"learning_rate": 4.984764998522357e-05,
|
| 121 |
+
"loss": 3.8839893341064453,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.10049363573019814,
|
| 126 |
+
"grad_norm": 22.662349700927734,
|
| 127 |
+
"learning_rate": 4.9826241921747555e-05,
|
| 128 |
+
"loss": 3.063496971130371,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.10677448796333552,
|
| 133 |
+
"grad_norm": 25.49854278564453,
|
| 134 |
+
"learning_rate": 4.9803432179971055e-05,
|
| 135 |
+
"loss": 3.0309188842773436,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.11305534019647291,
|
| 140 |
+
"grad_norm": 15.640332221984863,
|
| 141 |
+
"learning_rate": 4.9779222047721693e-05,
|
| 142 |
+
"loss": 2.6161094665527345,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1193361924296103,
|
| 147 |
+
"grad_norm": 13.35746955871582,
|
| 148 |
+
"learning_rate": 4.9753612891892514e-05,
|
| 149 |
+
"loss": 2.3498945236206055,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.12561704466274767,
|
| 154 |
+
"grad_norm": 10.242664337158203,
|
| 155 |
+
"learning_rate": 4.9726606158364765e-05,
|
| 156 |
+
"loss": 2.069006156921387,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.13189789689588505,
|
| 161 |
+
"grad_norm": 8.599258422851562,
|
| 162 |
+
"learning_rate": 4.969965664306835e-05,
|
| 163 |
+
"loss": 1.7923540115356444,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.13817874912902245,
|
| 168 |
+
"grad_norm": 5.654482364654541,
|
| 169 |
+
"learning_rate": 4.966992909049121e-05,
|
| 170 |
+
"loss": 1.7063348770141602,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.14445960136215982,
|
| 175 |
+
"grad_norm": 9.39411449432373,
|
| 176 |
+
"learning_rate": 4.9638808684964765e-05,
|
| 177 |
+
"loss": 1.4251695632934571,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.15074045359529722,
|
| 182 |
+
"grad_norm": 4.354184627532959,
|
| 183 |
+
"learning_rate": 4.960629718353295e-05,
|
| 184 |
+
"loss": 1.2388976097106934,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1570213058284346,
|
| 189 |
+
"grad_norm": 4.586623191833496,
|
| 190 |
+
"learning_rate": 4.9572396421780296e-05,
|
| 191 |
+
"loss": 1.2111310958862305,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.16330215806157197,
|
| 196 |
+
"grad_norm": 3.979524850845337,
|
| 197 |
+
"learning_rate": 4.9537108313728365e-05,
|
| 198 |
+
"loss": 1.4075250625610352,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.16958301029470937,
|
| 203 |
+
"grad_norm": 3.687380790710449,
|
| 204 |
+
"learning_rate": 4.9500434851727635e-05,
|
| 205 |
+
"loss": 0.9275534629821778,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.17586386252784675,
|
| 210 |
+
"grad_norm": 2.872728109359741,
|
| 211 |
+
"learning_rate": 4.946431376374067e-05,
|
| 212 |
+
"loss": 1.418515110015869,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.18214471476098412,
|
| 217 |
+
"grad_norm": 2.1888179779052734,
|
| 218 |
+
"learning_rate": 4.9424944888143395e-05,
|
| 219 |
+
"loss": 0.7708720684051513,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.18842556699412152,
|
| 224 |
+
"grad_norm": 2.96533465385437,
|
| 225 |
+
"learning_rate": 4.9384196991293205e-05,
|
| 226 |
+
"loss": 0.6360139369964599,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.1947064192272589,
|
| 231 |
+
"grad_norm": 3.389042615890503,
|
| 232 |
+
"learning_rate": 4.9342072373797774e-05,
|
| 233 |
+
"loss": 1.2306670188903808,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.20098727146039627,
|
| 238 |
+
"grad_norm": 3.054168939590454,
|
| 239 |
+
"learning_rate": 4.9298573413993806e-05,
|
| 240 |
+
"loss": 0.684608793258667,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.20726812369353367,
|
| 245 |
+
"grad_norm": 2.2155184745788574,
|
| 246 |
+
"learning_rate": 4.925370256781273e-05,
|
| 247 |
+
"loss": 0.6700978755950928,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.21354897592667105,
|
| 252 |
+
"grad_norm": 1.223708987236023,
|
| 253 |
+
"learning_rate": 4.920746236864208e-05,
|
| 254 |
+
"loss": 0.707273292541504,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.21982982815980842,
|
| 259 |
+
"grad_norm": 1.1626676321029663,
|
| 260 |
+
"learning_rate": 4.915985542718241e-05,
|
| 261 |
+
"loss": 0.6534492492675781,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.22611068039294582,
|
| 266 |
+
"grad_norm": 1.1415362358093262,
|
| 267 |
+
"learning_rate": 4.9110884431299935e-05,
|
| 268 |
+
"loss": 0.45153131484985354,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.2323915326260832,
|
| 273 |
+
"grad_norm": 1.0272735357284546,
|
| 274 |
+
"learning_rate": 4.906055214587476e-05,
|
| 275 |
+
"loss": 0.46799488067626954,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.2386723848592206,
|
| 280 |
+
"grad_norm": 1.2896318435668945,
|
| 281 |
+
"learning_rate": 4.9008861412644756e-05,
|
| 282 |
+
"loss": 0.4087998390197754,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.24495323709235797,
|
| 287 |
+
"grad_norm": 0.873556911945343,
|
| 288 |
+
"learning_rate": 4.895581515004517e-05,
|
| 289 |
+
"loss": 0.3405194044113159,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.25123408932549535,
|
| 294 |
+
"grad_norm": 1.2751230001449585,
|
| 295 |
+
"learning_rate": 4.89014163530438e-05,
|
| 296 |
+
"loss": 0.401216459274292,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.25751494155863275,
|
| 301 |
+
"grad_norm": 0.8582271337509155,
|
| 302 |
+
"learning_rate": 4.8845668092971934e-05,
|
| 303 |
+
"loss": 0.36631605625152586,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.2637957937917701,
|
| 308 |
+
"grad_norm": 0.7967858910560608,
|
| 309 |
+
"learning_rate": 4.87885735173509e-05,
|
| 310 |
+
"loss": 0.33444316387176515,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2700766460249075,
|
| 315 |
+
"grad_norm": 0.7824229001998901,
|
| 316 |
+
"learning_rate": 4.873013584971442e-05,
|
| 317 |
+
"loss": 0.3244049549102783,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.2763574982580449,
|
| 322 |
+
"grad_norm": 0.774782121181488,
|
| 323 |
+
"learning_rate": 4.867035838942657e-05,
|
| 324 |
+
"loss": 0.6262166023254394,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2826383504911823,
|
| 329 |
+
"grad_norm": 0.6737077832221985,
|
| 330 |
+
"learning_rate": 4.86092445114955e-05,
|
| 331 |
+
"loss": 0.28058154582977296,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.28891920272431965,
|
| 336 |
+
"grad_norm": 1.0165168046951294,
|
| 337 |
+
"learning_rate": 4.8546797666382896e-05,
|
| 338 |
+
"loss": 0.3938837289810181,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.29520005495745705,
|
| 343 |
+
"grad_norm": 1.093992829322815,
|
| 344 |
+
"learning_rate": 4.8483021379809154e-05,
|
| 345 |
+
"loss": 0.3142520904541016,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.30148090719059445,
|
| 350 |
+
"grad_norm": 0.6402747631072998,
|
| 351 |
+
"learning_rate": 4.8417919252554354e-05,
|
| 352 |
+
"loss": 0.29062786102294924,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3077617594237318,
|
| 357 |
+
"grad_norm": 1.9970321655273438,
|
| 358 |
+
"learning_rate": 4.8351494960254916e-05,
|
| 359 |
+
"loss": 0.31338260173797605,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.3140426116568692,
|
| 364 |
+
"grad_norm": 1.1727873086929321,
|
| 365 |
+
"learning_rate": 4.8283752253196094e-05,
|
| 366 |
+
"loss": 0.2333289623260498,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3203234638900066,
|
| 371 |
+
"grad_norm": 0.7795476913452148,
|
| 372 |
+
"learning_rate": 4.8214694956100245e-05,
|
| 373 |
+
"loss": 0.2488691806793213,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.32660431612314395,
|
| 378 |
+
"grad_norm": 1.1675890684127808,
|
| 379 |
+
"learning_rate": 4.814432696791089e-05,
|
| 380 |
+
"loss": 0.22975556850433348,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.33288516835628135,
|
| 385 |
+
"grad_norm": 0.5809735059738159,
|
| 386 |
+
"learning_rate": 4.807265226157255e-05,
|
| 387 |
+
"loss": 0.2238070011138916,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.33916602058941875,
|
| 392 |
+
"grad_norm": 0.6172874569892883,
|
| 393 |
+
"learning_rate": 4.7999674883806465e-05,
|
| 394 |
+
"loss": 0.22853121757507325,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3454468728225561,
|
| 399 |
+
"grad_norm": 0.9794221520423889,
|
| 400 |
+
"learning_rate": 4.792539895488213e-05,
|
| 401 |
+
"loss": 0.2538387060165405,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.3517277250556935,
|
| 406 |
+
"grad_norm": 0.6366945505142212,
|
| 407 |
+
"learning_rate": 4.784982866838462e-05,
|
| 408 |
+
"loss": 0.207783842086792,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.3580085772888309,
|
| 413 |
+
"grad_norm": 0.5059712529182434,
|
| 414 |
+
"learning_rate": 4.777296829097785e-05,
|
| 415 |
+
"loss": 0.2022031545639038,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.36428942952196824,
|
| 420 |
+
"grad_norm": 0.42327556014060974,
|
| 421 |
+
"learning_rate": 4.769482216216369e-05,
|
| 422 |
+
"loss": 0.19924311637878417,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.37057028175510565,
|
| 427 |
+
"grad_norm": 1.0261644124984741,
|
| 428 |
+
"learning_rate": 4.761539469403695e-05,
|
| 429 |
+
"loss": 0.1949863314628601,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.37685113398824305,
|
| 434 |
+
"grad_norm": 0.78790283203125,
|
| 435 |
+
"learning_rate": 4.7534690371036274e-05,
|
| 436 |
+
"loss": 0.1740255832672119,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3831319862213804,
|
| 441 |
+
"grad_norm": 0.49199551343917847,
|
| 442 |
+
"learning_rate": 4.7452713749690934e-05,
|
| 443 |
+
"loss": 0.16614840030670167,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.3894128384545178,
|
| 448 |
+
"grad_norm": 0.5265403985977173,
|
| 449 |
+
"learning_rate": 4.73694694583636e-05,
|
| 450 |
+
"loss": 0.17591538429260253,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.3956936906876552,
|
| 455 |
+
"grad_norm": 0.5221320390701294,
|
| 456 |
+
"learning_rate": 4.728496219698902e-05,
|
| 457 |
+
"loss": 0.18058695793151855,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.40197454292079254,
|
| 462 |
+
"grad_norm": 1.2317312955856323,
|
| 463 |
+
"learning_rate": 4.719919673680863e-05,
|
| 464 |
+
"loss": 0.12510637044906617,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.40825539515392995,
|
| 469 |
+
"grad_norm": 0.21159659326076508,
|
| 470 |
+
"learning_rate": 4.711217792010122e-05,
|
| 471 |
+
"loss": 0.3407147884368896,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.41453624738706735,
|
| 476 |
+
"grad_norm": 1.4082627296447754,
|
| 477 |
+
"learning_rate": 4.70239106599095e-05,
|
| 478 |
+
"loss": 0.16848219633102418,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.4208170996202047,
|
| 483 |
+
"grad_norm": 0.35336634516716003,
|
| 484 |
+
"learning_rate": 4.6934399939762746e-05,
|
| 485 |
+
"loss": 0.13055137395858765,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.4270979518533421,
|
| 490 |
+
"grad_norm": 0.8723010420799255,
|
| 491 |
+
"learning_rate": 4.684365081339541e-05,
|
| 492 |
+
"loss": 0.1356166958808899,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.4333788040864795,
|
| 497 |
+
"grad_norm": 0.49977484345436096,
|
| 498 |
+
"learning_rate": 4.675166840446179e-05,
|
| 499 |
+
"loss": 0.17129049301147461,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.43965965631961684,
|
| 504 |
+
"grad_norm": 0.8447715640068054,
|
| 505 |
+
"learning_rate": 4.6658457906246754e-05,
|
| 506 |
+
"loss": 0.16764267683029174,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.44594050855275424,
|
| 511 |
+
"grad_norm": 0.491344153881073,
|
| 512 |
+
"learning_rate": 4.656402458137255e-05,
|
| 513 |
+
"loss": 0.2734020709991455,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.45222136078589165,
|
| 518 |
+
"grad_norm": 0.7019853591918945,
|
| 519 |
+
"learning_rate": 4.646837376150164e-05,
|
| 520 |
+
"loss": 0.15193712711334229,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.458502213019029,
|
| 525 |
+
"grad_norm": 0.5096240639686584,
|
| 526 |
+
"learning_rate": 4.637151084703571e-05,
|
| 527 |
+
"loss": 0.31453657150268555,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.4647830652521664,
|
| 532 |
+
"grad_norm": 0.457865446805954,
|
| 533 |
+
"learning_rate": 4.6273441306810724e-05,
|
| 534 |
+
"loss": 0.14785597324371338,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.4710639174853038,
|
| 539 |
+
"grad_norm": 0.5041202306747437,
|
| 540 |
+
"learning_rate": 4.617417067778825e-05,
|
| 541 |
+
"loss": 0.13907097578048705,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.4773447697184412,
|
| 546 |
+
"grad_norm": 3.153047800064087,
|
| 547 |
+
"learning_rate": 4.6073704564742734e-05,
|
| 548 |
+
"loss": 0.1594400644302368,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.48362562195157854,
|
| 553 |
+
"grad_norm": 6.409017562866211,
|
| 554 |
+
"learning_rate": 4.5972048639945105e-05,
|
| 555 |
+
"loss": 0.1708819031715393,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.48990647418471595,
|
| 560 |
+
"grad_norm": 0.5795732736587524,
|
| 561 |
+
"learning_rate": 4.5869208642842536e-05,
|
| 562 |
+
"loss": 0.5376199245452881,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.49618732641785335,
|
| 567 |
+
"grad_norm": 1.402137279510498,
|
| 568 |
+
"learning_rate": 4.5765190379734384e-05,
|
| 569 |
+
"loss": 0.13104203939437867,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.5024681786509907,
|
| 574 |
+
"grad_norm": 1.01236093044281,
|
| 575 |
+
"learning_rate": 4.565999972344436e-05,
|
| 576 |
+
"loss": 0.14823287725448608,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.508749030884128,
|
| 581 |
+
"grad_norm": 0.47942015528678894,
|
| 582 |
+
"learning_rate": 4.5553642612988954e-05,
|
| 583 |
+
"loss": 0.13399872779846192,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.5150298831172655,
|
| 588 |
+
"grad_norm": 0.5736086964607239,
|
| 589 |
+
"learning_rate": 4.544612505324213e-05,
|
| 590 |
+
"loss": 0.09690673351287842,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.5213107353504028,
|
| 595 |
+
"grad_norm": 0.4799337387084961,
|
| 596 |
+
"learning_rate": 4.533745311459631e-05,
|
| 597 |
+
"loss": 0.12333678007125855,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.5275915875835402,
|
| 602 |
+
"grad_norm": 0.49539658427238464,
|
| 603 |
+
"learning_rate": 4.5227632932619604e-05,
|
| 604 |
+
"loss": 0.1237714409828186,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.5338724398166776,
|
| 609 |
+
"grad_norm": 3.5184292793273926,
|
| 610 |
+
"learning_rate": 4.5116670707709396e-05,
|
| 611 |
+
"loss": 0.12902543544769288,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.540153292049815,
|
| 616 |
+
"grad_norm": 0.543497622013092,
|
| 617 |
+
"learning_rate": 4.5004572704742345e-05,
|
| 618 |
+
"loss": 0.11819875240325928,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.5464341442829523,
|
| 623 |
+
"grad_norm": 0.47288012504577637,
|
| 624 |
+
"learning_rate": 4.489134525272057e-05,
|
| 625 |
+
"loss": 0.09024000763893128,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.5527149965160898,
|
| 630 |
+
"grad_norm": 0.5600817203521729,
|
| 631 |
+
"learning_rate": 4.477699474441441e-05,
|
| 632 |
+
"loss": 0.10022807121276855,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.5589958487492271,
|
| 637 |
+
"grad_norm": 0.38338902592658997,
|
| 638 |
+
"learning_rate": 4.466152763600142e-05,
|
| 639 |
+
"loss": 0.1357239603996277,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.5652767009823646,
|
| 644 |
+
"grad_norm": 0.12702441215515137,
|
| 645 |
+
"learning_rate": 4.454495044670189e-05,
|
| 646 |
+
"loss": 0.36903703212738037,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.571557553215502,
|
| 651 |
+
"grad_norm": 0.32971400022506714,
|
| 652 |
+
"learning_rate": 4.4427269758410806e-05,
|
| 653 |
+
"loss": 0.09078516364097595,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5778384054486393,
|
| 658 |
+
"grad_norm": 0.5972228050231934,
|
| 659 |
+
"learning_rate": 4.4308492215326145e-05,
|
| 660 |
+
"loss": 0.13515870571136473,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5841192576817767,
|
| 665 |
+
"grad_norm": 0.4297616183757782,
|
| 666 |
+
"learning_rate": 4.418862452357387e-05,
|
| 667 |
+
"loss": 0.16306908130645753,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.5904001099149141,
|
| 672 |
+
"grad_norm": 0.3069750666618347,
|
| 673 |
+
"learning_rate": 4.4067673450829184e-05,
|
| 674 |
+
"loss": 0.11015008687973023,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.5966809621480514,
|
| 679 |
+
"grad_norm": 0.2897976040840149,
|
| 680 |
+
"learning_rate": 4.394564582593452e-05,
|
| 681 |
+
"loss": 0.10074495077133179,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.6029618143811889,
|
| 686 |
+
"grad_norm": 0.2866128981113434,
|
| 687 |
+
"learning_rate": 4.382254853851396e-05,
|
| 688 |
+
"loss": 0.10123426914215088,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.6092426666143262,
|
| 693 |
+
"grad_norm": 0.611520528793335,
|
| 694 |
+
"learning_rate": 4.3698388538584214e-05,
|
| 695 |
+
"loss": 0.1677260398864746,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.6155235188474636,
|
| 700 |
+
"grad_norm": 0.6373476982116699,
|
| 701 |
+
"learning_rate": 4.357317283616228e-05,
|
| 702 |
+
"loss": 0.07637218236923218,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.621804371080601,
|
| 707 |
+
"grad_norm": 0.24775440990924835,
|
| 708 |
+
"learning_rate": 4.3446908500869624e-05,
|
| 709 |
+
"loss": 0.07576732635498047,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.6280852233137384,
|
| 714 |
+
"grad_norm": 0.36698204278945923,
|
| 715 |
+
"learning_rate": 4.3319602661533054e-05,
|
| 716 |
+
"loss": 0.08246396780014038,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.6343660755468757,
|
| 721 |
+
"grad_norm": 0.4456537365913391,
|
| 722 |
+
"learning_rate": 4.319126250578223e-05,
|
| 723 |
+
"loss": 0.08348861336708069,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.6406469277800132,
|
| 728 |
+
"grad_norm": 0.12586534023284912,
|
| 729 |
+
"learning_rate": 4.306189527964382e-05,
|
| 730 |
+
"loss": 0.11425427198410035,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.6469277800131505,
|
| 735 |
+
"grad_norm": 0.42404818534851074,
|
| 736 |
+
"learning_rate": 4.2931508287132435e-05,
|
| 737 |
+
"loss": 0.43116607666015627,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6532086322462879,
|
| 742 |
+
"grad_norm": 0.322374165058136,
|
| 743 |
+
"learning_rate": 4.2800108889838244e-05,
|
| 744 |
+
"loss": 0.07275817394256592,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.6594894844794253,
|
| 749 |
+
"grad_norm": 0.24643594026565552,
|
| 750 |
+
"learning_rate": 4.26677045065113e-05,
|
| 751 |
+
"loss": 0.0800868034362793,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6657703367125627,
|
| 756 |
+
"grad_norm": 0.32612672448158264,
|
| 757 |
+
"learning_rate": 4.253430261264272e-05,
|
| 758 |
+
"loss": 0.11347514390945435,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6720511889457,
|
| 763 |
+
"grad_norm": 0.9363917708396912,
|
| 764 |
+
"learning_rate": 4.239991074004262e-05,
|
| 765 |
+
"loss": 0.14267417192459106,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6783320411788375,
|
| 770 |
+
"grad_norm": 5.144097328186035,
|
| 771 |
+
"learning_rate": 4.226453647641484e-05,
|
| 772 |
+
"loss": 0.2039745807647705,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.6846128934119748,
|
| 777 |
+
"grad_norm": 0.275455117225647,
|
| 778 |
+
"learning_rate": 4.2128187464928604e-05,
|
| 779 |
+
"loss": 0.12193784713745118,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.6908937456451122,
|
| 784 |
+
"grad_norm": 0.23535878956317902,
|
| 785 |
+
"learning_rate": 4.1990871403786934e-05,
|
| 786 |
+
"loss": 0.08345263600349426,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.6971745978782496,
|
| 791 |
+
"grad_norm": 0.24043764173984528,
|
| 792 |
+
"learning_rate": 4.185259604579203e-05,
|
| 793 |
+
"loss": 0.06981859803199768,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.703455450111387,
|
| 798 |
+
"grad_norm": 0.6327846050262451,
|
| 799 |
+
"learning_rate": 4.171336919790759e-05,
|
| 800 |
+
"loss": 0.1702319025993347,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.7097363023445243,
|
| 805 |
+
"grad_norm": 0.059428345412015915,
|
| 806 |
+
"learning_rate": 4.1573198720817955e-05,
|
| 807 |
+
"loss": 0.07783352732658386,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.7160171545776618,
|
| 812 |
+
"grad_norm": 0.4627244770526886,
|
| 813 |
+
"learning_rate": 4.143209252848433e-05,
|
| 814 |
+
"loss": 0.07844600677490235,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.7222980068107991,
|
| 819 |
+
"grad_norm": 0.7847932577133179,
|
| 820 |
+
"learning_rate": 4.1290058587698054e-05,
|
| 821 |
+
"loss": 0.09436154365539551,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.7285788590439365,
|
| 826 |
+
"grad_norm": 0.6646613478660583,
|
| 827 |
+
"learning_rate": 4.114710491763063e-05,
|
| 828 |
+
"loss": 0.10847301483154297,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.734859711277074,
|
| 833 |
+
"grad_norm": 0.13904166221618652,
|
| 834 |
+
"learning_rate": 4.100323958938112e-05,
|
| 835 |
+
"loss": 0.10292520523071289,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.7411405635102113,
|
| 840 |
+
"grad_norm": 0.17684051394462585,
|
| 841 |
+
"learning_rate": 4.085847072552036e-05,
|
| 842 |
+
"loss": 0.10409547090530395,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.7474214157433486,
|
| 847 |
+
"grad_norm": 2.33097767829895,
|
| 848 |
+
"learning_rate": 4.0712806499632397e-05,
|
| 849 |
+
"loss": 0.07961801886558532,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.7537022679764861,
|
| 854 |
+
"grad_norm": 0.476825475692749,
|
| 855 |
+
"learning_rate": 4.0566255135853014e-05,
|
| 856 |
+
"loss": 0.07112305760383605,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.7599831202096234,
|
| 861 |
+
"grad_norm": 0.5014712810516357,
|
| 862 |
+
"learning_rate": 4.04188249084054e-05,
|
| 863 |
+
"loss": 0.1152498722076416,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.7662639724427608,
|
| 868 |
+
"grad_norm": 0.4747202694416046,
|
| 869 |
+
"learning_rate": 4.0270524141133e-05,
|
| 870 |
+
"loss": 0.12496150732040405,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.7725448246758982,
|
| 875 |
+
"grad_norm": 0.3705868721008301,
|
| 876 |
+
"learning_rate": 4.01213612070295e-05,
|
| 877 |
+
"loss": 0.09812939167022705,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.7788256769090356,
|
| 882 |
+
"grad_norm": 0.18966151773929596,
|
| 883 |
+
"learning_rate": 3.997134452776617e-05,
|
| 884 |
+
"loss": 0.09260130524635315,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.7851065291421729,
|
| 889 |
+
"grad_norm": 0.4178507328033447,
|
| 890 |
+
"learning_rate": 3.982048257321631e-05,
|
| 891 |
+
"loss": 0.08796253800392151,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.7913873813753104,
|
| 896 |
+
"grad_norm": 4.684873580932617,
|
| 897 |
+
"learning_rate": 3.966878386097711e-05,
|
| 898 |
+
"loss": 0.09675108194351197,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.7976682336084477,
|
| 903 |
+
"grad_norm": 0.7546767592430115,
|
| 904 |
+
"learning_rate": 3.95162569558887e-05,
|
| 905 |
+
"loss": 0.061058801412582395,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.8039490858415851,
|
| 910 |
+
"grad_norm": 1.2984436750411987,
|
| 911 |
+
"learning_rate": 3.936291046955061e-05,
|
| 912 |
+
"loss": 0.0732529878616333,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.8102299380747225,
|
| 917 |
+
"grad_norm": 0.08046652376651764,
|
| 918 |
+
"learning_rate": 3.920875305983553e-05,
|
| 919 |
+
"loss": 0.056511855125427245,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.8165107903078599,
|
| 924 |
+
"grad_norm": 0.24541246891021729,
|
| 925 |
+
"learning_rate": 3.905379343040054e-05,
|
| 926 |
+
"loss": 0.06492850780487061,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.8227916425409972,
|
| 931 |
+
"grad_norm": 0.34249547123908997,
|
| 932 |
+
"learning_rate": 3.889804033019567e-05,
|
| 933 |
+
"loss": 0.06372414827346802,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.8290724947741347,
|
| 938 |
+
"grad_norm": 0.6350385546684265,
|
| 939 |
+
"learning_rate": 3.8741502552969956e-05,
|
| 940 |
+
"loss": 0.07334259748458863,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.835353347007272,
|
| 945 |
+
"grad_norm": 0.5676931738853455,
|
| 946 |
+
"learning_rate": 3.8584188936774925e-05,
|
| 947 |
+
"loss": 0.051751017570495605,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.8416341992404094,
|
| 952 |
+
"grad_norm": 1.5374674797058105,
|
| 953 |
+
"learning_rate": 3.842610836346563e-05,
|
| 954 |
+
"loss": 0.13946666717529296,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.8479150514735468,
|
| 959 |
+
"grad_norm": 0.4549136459827423,
|
| 960 |
+
"learning_rate": 3.8267269758199175e-05,
|
| 961 |
+
"loss": 0.06678192019462585,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.8541959037066842,
|
| 966 |
+
"grad_norm": 0.4328550696372986,
|
| 967 |
+
"learning_rate": 3.8107682088930794e-05,
|
| 968 |
+
"loss": 0.0756498098373413,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.8604767559398215,
|
| 973 |
+
"grad_norm": 1.7214152812957764,
|
| 974 |
+
"learning_rate": 3.7947354365907536e-05,
|
| 975 |
+
"loss": 0.07417964935302734,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.866757608172959,
|
| 980 |
+
"grad_norm": 0.3970869481563568,
|
| 981 |
+
"learning_rate": 3.7786295641159544e-05,
|
| 982 |
+
"loss": 0.055640166997909545,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.8730384604060963,
|
| 987 |
+
"grad_norm": 0.16465847194194794,
|
| 988 |
+
"learning_rate": 3.7624515007989004e-05,
|
| 989 |
+
"loss": 0.057208091020584106,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.8793193126392337,
|
| 994 |
+
"grad_norm": 1.0308940410614014,
|
| 995 |
+
"learning_rate": 3.746202160045669e-05,
|
| 996 |
+
"loss": 0.1124604344367981,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.8856001648723711,
|
| 1001 |
+
"grad_norm": 1.095229983329773,
|
| 1002 |
+
"learning_rate": 3.729882459286632e-05,
|
| 1003 |
+
"loss": 0.06749147176742554,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.8918810171055085,
|
| 1008 |
+
"grad_norm": 0.7268246412277222,
|
| 1009 |
+
"learning_rate": 3.713493319924655e-05,
|
| 1010 |
+
"loss": 0.05816763639450073,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.8981618693386458,
|
| 1015 |
+
"grad_norm": 0.06477148085832596,
|
| 1016 |
+
"learning_rate": 3.697035667283072e-05,
|
| 1017 |
+
"loss": 0.07365052700042725,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.9044427215717833,
|
| 1022 |
+
"grad_norm": 0.14640307426452637,
|
| 1023 |
+
"learning_rate": 3.68051043055345e-05,
|
| 1024 |
+
"loss": 0.10839139223098755,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.9107235738049206,
|
| 1029 |
+
"grad_norm": 0.6694417595863342,
|
| 1030 |
+
"learning_rate": 3.66391854274312e-05,
|
| 1031 |
+
"loss": 0.0827883541584015,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.917004426038058,
|
| 1036 |
+
"grad_norm": 0.7503764629364014,
|
| 1037 |
+
"learning_rate": 3.647260940622504e-05,
|
| 1038 |
+
"loss": 0.05839287042617798,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.9232852782711954,
|
| 1043 |
+
"grad_norm": 0.6776005029678345,
|
| 1044 |
+
"learning_rate": 3.6305385646722224e-05,
|
| 1045 |
+
"loss": 0.08498733639717101,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.9295661305043328,
|
| 1050 |
+
"grad_norm": 0.1537044644355774,
|
| 1051 |
+
"learning_rate": 3.613752359029998e-05,
|
| 1052 |
+
"loss": 0.1408205509185791,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.9358469827374702,
|
| 1057 |
+
"grad_norm": 0.18748170137405396,
|
| 1058 |
+
"learning_rate": 3.5969032714373495e-05,
|
| 1059 |
+
"loss": 0.07079569697380066,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.9421278349706076,
|
| 1064 |
+
"grad_norm": 1.4993492364883423,
|
| 1065 |
+
"learning_rate": 3.57999225318608e-05,
|
| 1066 |
+
"loss": 0.07452060580253601,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.9484086872037449,
|
| 1071 |
+
"grad_norm": 2.759636163711548,
|
| 1072 |
+
"learning_rate": 3.5638702921686935e-05,
|
| 1073 |
+
"loss": 0.30155444145202637,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.9546895394368824,
|
| 1078 |
+
"grad_norm": 0.35174673795700073,
|
| 1079 |
+
"learning_rate": 3.546841258479434e-05,
|
| 1080 |
+
"loss": 0.07043100595474243,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.9609703916700197,
|
| 1085 |
+
"grad_norm": 0.1061045303940773,
|
| 1086 |
+
"learning_rate": 3.529753120609982e-05,
|
| 1087 |
+
"loss": 0.06520496010780334,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.9672512439031571,
|
| 1092 |
+
"grad_norm": 0.32999858260154724,
|
| 1093 |
+
"learning_rate": 3.51260684334881e-05,
|
| 1094 |
+
"loss": 0.11166098117828369,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.9735320961362945,
|
| 1099 |
+
"grad_norm": 0.527561366558075,
|
| 1100 |
+
"learning_rate": 3.4954033947669163e-05,
|
| 1101 |
+
"loss": 0.05884957313537598,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.9798129483694319,
|
| 1106 |
+
"grad_norm": 0.21894922852516174,
|
| 1107 |
+
"learning_rate": 3.478143746163164e-05,
|
| 1108 |
+
"loss": 0.04760840237140655,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9860938006025692,
|
| 1113 |
+
"grad_norm": 0.39011839032173157,
|
| 1114 |
+
"learning_rate": 3.460828872009444e-05,
|
| 1115 |
+
"loss": 0.07473071217536927,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.9923746528357067,
|
| 1120 |
+
"grad_norm": 0.24014221131801605,
|
| 1121 |
+
"learning_rate": 3.443459749895658e-05,
|
| 1122 |
+
"loss": 0.05981050133705139,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.998655505068844,
|
| 1127 |
+
"grad_norm": 0.4369584918022156,
|
| 1128 |
+
"learning_rate": 3.4260373604745205e-05,
|
| 1129 |
+
"loss": 0.06494853496551514,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.004710639174853,
|
| 1134 |
+
"grad_norm": 0.11435407400131226,
|
| 1135 |
+
"learning_rate": 3.4085626874061974e-05,
|
| 1136 |
+
"loss": 0.037298232316970825,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.0109914914079905,
|
| 1141 |
+
"grad_norm": 0.10405384004116058,
|
| 1142 |
+
"learning_rate": 3.391036717302761e-05,
|
| 1143 |
+
"loss": 0.03180496394634247,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.0172723436411277,
|
| 1148 |
+
"grad_norm": 0.2586648166179657,
|
| 1149 |
+
"learning_rate": 3.373460439672495e-05,
|
| 1150 |
+
"loss": 0.04955384433269501,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.0235531958742652,
|
| 1155 |
+
"grad_norm": 0.08358602225780487,
|
| 1156 |
+
"learning_rate": 3.35583484686402e-05,
|
| 1157 |
+
"loss": 0.05318580269813537,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.0298340481074026,
|
| 1162 |
+
"grad_norm": 0.07553794234991074,
|
| 1163 |
+
"learning_rate": 3.338160934010271e-05,
|
| 1164 |
+
"loss": 0.08674532771110535,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.0361149003405399,
|
| 1169 |
+
"grad_norm": 0.2271793782711029,
|
| 1170 |
+
"learning_rate": 3.32043969897231e-05,
|
| 1171 |
+
"loss": 0.04139872789382935,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.0423957525736773,
|
| 1176 |
+
"grad_norm": 0.25413981080055237,
|
| 1177 |
+
"learning_rate": 3.302672142282988e-05,
|
| 1178 |
+
"loss": 0.035733970999717715,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.0486766048068148,
|
| 1183 |
+
"grad_norm": 0.09163467586040497,
|
| 1184 |
+
"learning_rate": 3.284859267090456e-05,
|
| 1185 |
+
"loss": 0.08379928469657898,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.054957457039952,
|
| 1190 |
+
"grad_norm": 0.07256153970956802,
|
| 1191 |
+
"learning_rate": 3.267002079101525e-05,
|
| 1192 |
+
"loss": 0.03688819110393524,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.0612383092730895,
|
| 1197 |
+
"grad_norm": 0.305923193693161,
|
| 1198 |
+
"learning_rate": 3.2491015865248884e-05,
|
| 1199 |
+
"loss": 0.049816560745239255,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.067519161506227,
|
| 1204 |
+
"grad_norm": 0.046878550201654434,
|
| 1205 |
+
"learning_rate": 3.231158800014197e-05,
|
| 1206 |
+
"loss": 0.03804933726787567,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.0738000137393642,
|
| 1211 |
+
"grad_norm": 0.6410548686981201,
|
| 1212 |
+
"learning_rate": 3.213174732610995e-05,
|
| 1213 |
+
"loss": 0.0333125501871109,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.0800808659725016,
|
| 1218 |
+
"grad_norm": 0.06022074818611145,
|
| 1219 |
+
"learning_rate": 3.195150399687531e-05,
|
| 1220 |
+
"loss": 0.028510311245918275,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.086361718205639,
|
| 1225 |
+
"grad_norm": 0.41786953806877136,
|
| 1226 |
+
"learning_rate": 3.177086818889423e-05,
|
| 1227 |
+
"loss": 0.046647512912750246,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.0926425704387763,
|
| 1232 |
+
"grad_norm": 0.1293671578168869,
|
| 1233 |
+
"learning_rate": 3.1589850100782085e-05,
|
| 1234 |
+
"loss": 0.03643852770328522,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.0989234226719138,
|
| 1239 |
+
"grad_norm": 0.05348878353834152,
|
| 1240 |
+
"learning_rate": 3.1408459952737594e-05,
|
| 1241 |
+
"loss": 0.03161140382289886,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.1052042749050512,
|
| 1246 |
+
"grad_norm": 0.29638731479644775,
|
| 1247 |
+
"learning_rate": 3.122670798596583e-05,
|
| 1248 |
+
"loss": 0.04662284553050995,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.1114851271381885,
|
| 1253 |
+
"grad_norm": 0.11425400525331497,
|
| 1254 |
+
"learning_rate": 3.1044604462099965e-05,
|
| 1255 |
+
"loss": 0.04306844174861908,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.117765979371326,
|
| 1260 |
+
"grad_norm": 13.269792556762695,
|
| 1261 |
+
"learning_rate": 3.0862159662621933e-05,
|
| 1262 |
+
"loss": 0.10026944875717163,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.1240468316044634,
|
| 1267 |
+
"grad_norm": 0.09814783930778503,
|
| 1268 |
+
"learning_rate": 3.067938388828195e-05,
|
| 1269 |
+
"loss": 0.03279039561748505,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.1303276838376006,
|
| 1274 |
+
"grad_norm": 2.8222439289093018,
|
| 1275 |
+
"learning_rate": 3.049628745851689e-05,
|
| 1276 |
+
"loss": 0.03593765795230865,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.136608536070738,
|
| 1281 |
+
"grad_norm": 0.14487864077091217,
|
| 1282 |
+
"learning_rate": 3.031288071086772e-05,
|
| 1283 |
+
"loss": 0.0726437509059906,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.1428893883038755,
|
| 1288 |
+
"grad_norm": 1.1210272312164307,
|
| 1289 |
+
"learning_rate": 3.0129174000395776e-05,
|
| 1290 |
+
"loss": 0.03795652091503143,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.1491702405370128,
|
| 1295 |
+
"grad_norm": 0.42077013850212097,
|
| 1296 |
+
"learning_rate": 2.99451776990982e-05,
|
| 1297 |
+
"loss": 0.04205833375453949,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.1554510927701502,
|
| 1302 |
+
"grad_norm": 0.08231453597545624,
|
| 1303 |
+
"learning_rate": 2.976090219532226e-05,
|
| 1304 |
+
"loss": 0.04487053751945495,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.1617319450032877,
|
| 1309 |
+
"grad_norm": 0.1877751648426056,
|
| 1310 |
+
"learning_rate": 2.9576357893178885e-05,
|
| 1311 |
+
"loss": 0.034676852822303775,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.168012797236425,
|
| 1316 |
+
"grad_norm": 0.6968119144439697,
|
| 1317 |
+
"learning_rate": 2.939155521195526e-05,
|
| 1318 |
+
"loss": 0.03220001757144928,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.1742936494695624,
|
| 1323 |
+
"grad_norm": 0.06414724141359329,
|
| 1324 |
+
"learning_rate": 2.9206504585526493e-05,
|
| 1325 |
+
"loss": 0.041881483793258664,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.1805745017026998,
|
| 1330 |
+
"grad_norm": 0.10654035210609436,
|
| 1331 |
+
"learning_rate": 2.9021216461766604e-05,
|
| 1332 |
+
"loss": 0.03662925064563751,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.186855353935837,
|
| 1337 |
+
"grad_norm": 0.04183540865778923,
|
| 1338 |
+
"learning_rate": 2.8835701301958574e-05,
|
| 1339 |
+
"loss": 0.04763528108596802,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.1931362061689745,
|
| 1344 |
+
"grad_norm": 2.18105411529541,
|
| 1345 |
+
"learning_rate": 2.864996958020375e-05,
|
| 1346 |
+
"loss": 0.06307411193847656,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.199417058402112,
|
| 1351 |
+
"grad_norm": 0.3592161536216736,
|
| 1352 |
+
"learning_rate": 2.8464031782830474e-05,
|
| 1353 |
+
"loss": 0.051524901390075685,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.2056979106352492,
|
| 1358 |
+
"grad_norm": 0.11931532621383667,
|
| 1359 |
+
"learning_rate": 2.827789840780201e-05,
|
| 1360 |
+
"loss": 0.035766962170600894,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.2119787628683867,
|
| 1365 |
+
"grad_norm": 0.06202741712331772,
|
| 1366 |
+
"learning_rate": 2.8091579964123854e-05,
|
| 1367 |
+
"loss": 0.04704097211360932,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.2182596151015241,
|
| 1372 |
+
"grad_norm": 0.04776179790496826,
|
| 1373 |
+
"learning_rate": 2.7905086971250393e-05,
|
| 1374 |
+
"loss": 0.028332751989364625,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.2245404673346614,
|
| 1379 |
+
"grad_norm": 0.2371724396944046,
|
| 1380 |
+
"learning_rate": 2.771842995849098e-05,
|
| 1381 |
+
"loss": 0.027332887053489685,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.2308213195677988,
|
| 1386 |
+
"grad_norm": 0.05050333961844444,
|
| 1387 |
+
"learning_rate": 2.753161946441546e-05,
|
| 1388 |
+
"loss": 0.02915753722190857,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.2371021718009363,
|
| 1393 |
+
"grad_norm": 0.08952230215072632,
|
| 1394 |
+
"learning_rate": 2.734466603625915e-05,
|
| 1395 |
+
"loss": 0.025665390491485595,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.2433830240340735,
|
| 1400 |
+
"grad_norm": 0.20825526118278503,
|
| 1401 |
+
"learning_rate": 2.7157580229327385e-05,
|
| 1402 |
+
"loss": 0.2276303768157959,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.249663876267211,
|
| 1407 |
+
"grad_norm": 1.8631341457366943,
|
| 1408 |
+
"learning_rate": 2.6970372606399525e-05,
|
| 1409 |
+
"loss": 0.03601144254207611,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.2559447285003484,
|
| 1414 |
+
"grad_norm": 0.21078747510910034,
|
| 1415 |
+
"learning_rate": 2.6783053737132607e-05,
|
| 1416 |
+
"loss": 0.08526483774185181,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.2622255807334857,
|
| 1421 |
+
"grad_norm": 0.051980458199977875,
|
| 1422 |
+
"learning_rate": 2.6595634197464596e-05,
|
| 1423 |
+
"loss": 0.022251833975315095,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.2685064329666231,
|
| 1428 |
+
"grad_norm": 0.3019862771034241,
|
| 1429 |
+
"learning_rate": 2.6408124569017257e-05,
|
| 1430 |
+
"loss": 0.05009118318557739,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.2747872851997606,
|
| 1435 |
+
"grad_norm": 0.6371468305587769,
|
| 1436 |
+
"learning_rate": 2.622053543849872e-05,
|
| 1437 |
+
"loss": 0.040181088447570804,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.2810681374328978,
|
| 1442 |
+
"grad_norm": 1.1055502891540527,
|
| 1443 |
+
"learning_rate": 2.6032877397105783e-05,
|
| 1444 |
+
"loss": 0.02929992079734802,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.2873489896660353,
|
| 1449 |
+
"grad_norm": 0.26957738399505615,
|
| 1450 |
+
"learning_rate": 2.58451610399259e-05,
|
| 1451 |
+
"loss": 0.020054836571216584,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.2936298418991727,
|
| 1456 |
+
"grad_norm": 0.07279884070158005,
|
| 1457 |
+
"learning_rate": 2.5657396965339003e-05,
|
| 1458 |
+
"loss": 0.039886316657066344,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.29991069413231,
|
| 1463 |
+
"grad_norm": 0.09953224658966064,
|
| 1464 |
+
"learning_rate": 2.5469595774419158e-05,
|
| 1465 |
+
"loss": 0.03994681239128113,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.3061915463654474,
|
| 1470 |
+
"grad_norm": 0.14353635907173157,
|
| 1471 |
+
"learning_rate": 2.5281768070335988e-05,
|
| 1472 |
+
"loss": 0.03711488544940948,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.3124723985985849,
|
| 1477 |
+
"grad_norm": 0.27422836422920227,
|
| 1478 |
+
"learning_rate": 2.5093924457756018e-05,
|
| 1479 |
+
"loss": 0.023467497527599336,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.3187532508317221,
|
| 1484 |
+
"grad_norm": 0.0728367418050766,
|
| 1485 |
+
"learning_rate": 2.490607554224399e-05,
|
| 1486 |
+
"loss": 0.018936164677143097,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.3250341030648596,
|
| 1491 |
+
"grad_norm": 0.24521099030971527,
|
| 1492 |
+
"learning_rate": 2.4718231929664025e-05,
|
| 1493 |
+
"loss": 0.03874744176864624,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.331314955297997,
|
| 1498 |
+
"grad_norm": 0.13463203608989716,
|
| 1499 |
+
"learning_rate": 2.453040422558084e-05,
|
| 1500 |
+
"loss": 0.03414709270000458,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.3375958075311343,
|
| 1505 |
+
"grad_norm": 0.12021703273057938,
|
| 1506 |
+
"learning_rate": 2.4342603034661003e-05,
|
| 1507 |
+
"loss": 0.028204512596130372,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.3438766597642717,
|
| 1512 |
+
"grad_norm": 0.09300895035266876,
|
| 1513 |
+
"learning_rate": 2.4154838960074117e-05,
|
| 1514 |
+
"loss": 0.03560083508491516,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.3501575119974092,
|
| 1519 |
+
"grad_norm": 0.07003216445446014,
|
| 1520 |
+
"learning_rate": 2.3967122602894223e-05,
|
| 1521 |
+
"loss": 0.04477728307247162,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.3564383642305464,
|
| 1526 |
+
"grad_norm": 0.15798458456993103,
|
| 1527 |
+
"learning_rate": 2.377946456150128e-05,
|
| 1528 |
+
"loss": 0.05230234265327453,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.3627192164636839,
|
| 1533 |
+
"grad_norm": 0.05332406610250473,
|
| 1534 |
+
"learning_rate": 2.3591875430982752e-05,
|
| 1535 |
+
"loss": 0.029861843585968016,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.3690000686968213,
|
| 1540 |
+
"grad_norm": 0.05614598095417023,
|
| 1541 |
+
"learning_rate": 2.3404365802535413e-05,
|
| 1542 |
+
"loss": 0.03531986773014069,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.3752809209299586,
|
| 1547 |
+
"grad_norm": 0.33550721406936646,
|
| 1548 |
+
"learning_rate": 2.3216946262867396e-05,
|
| 1549 |
+
"loss": 0.031084492802619934,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.381561773163096,
|
| 1554 |
+
"grad_norm": 0.3401692807674408,
|
| 1555 |
+
"learning_rate": 2.3029627393600484e-05,
|
| 1556 |
+
"loss": 0.02833726108074188,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.3878426253962335,
|
| 1561 |
+
"grad_norm": 0.15662887692451477,
|
| 1562 |
+
"learning_rate": 2.2842419770672624e-05,
|
| 1563 |
+
"loss": 0.03976942598819733,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.3941234776293707,
|
| 1568 |
+
"grad_norm": 0.0530722439289093,
|
| 1569 |
+
"learning_rate": 2.265533396374085e-05,
|
| 1570 |
+
"loss": 0.03581877648830414,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.4004043298625082,
|
| 1575 |
+
"grad_norm": 0.09495271742343903,
|
| 1576 |
+
"learning_rate": 2.246838053558455e-05,
|
| 1577 |
+
"loss": 0.02298358678817749,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.4066851820956456,
|
| 1582 |
+
"grad_norm": 0.04610360413789749,
|
| 1583 |
+
"learning_rate": 2.2281570041509024e-05,
|
| 1584 |
+
"loss": 0.04205411374568939,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.4129660343287829,
|
| 1589 |
+
"grad_norm": 0.07970083504915237,
|
| 1590 |
+
"learning_rate": 2.209491302874961e-05,
|
| 1591 |
+
"loss": 0.028789132833480835,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.4192468865619203,
|
| 1596 |
+
"grad_norm": 1.6510498523712158,
|
| 1597 |
+
"learning_rate": 2.1908420035876148e-05,
|
| 1598 |
+
"loss": 0.030281484127044678,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.4255277387950578,
|
| 1603 |
+
"grad_norm": 0.061110883951187134,
|
| 1604 |
+
"learning_rate": 2.1722101592197995e-05,
|
| 1605 |
+
"loss": 0.026464346051216125,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.431808591028195,
|
| 1610 |
+
"grad_norm": 0.10222575813531876,
|
| 1611 |
+
"learning_rate": 2.1535968217169535e-05,
|
| 1612 |
+
"loss": 0.024321816861629486,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.4380894432613325,
|
| 1617 |
+
"grad_norm": 0.13632528483867645,
|
| 1618 |
+
"learning_rate": 2.135003041979625e-05,
|
| 1619 |
+
"loss": 0.022314466536045074,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.44437029549447,
|
| 1624 |
+
"grad_norm": 0.3248070478439331,
|
| 1625 |
+
"learning_rate": 2.1164298698041432e-05,
|
| 1626 |
+
"loss": 0.05798411965370178,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.4506511477276072,
|
| 1631 |
+
"grad_norm": 0.12926240265369415,
|
| 1632 |
+
"learning_rate": 2.0978783538233405e-05,
|
| 1633 |
+
"loss": 0.02899252772331238,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.4569319999607446,
|
| 1638 |
+
"grad_norm": 0.23640146851539612,
|
| 1639 |
+
"learning_rate": 2.0793495414473506e-05,
|
| 1640 |
+
"loss": 0.020399539172649382,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.463212852193882,
|
| 1645 |
+
"grad_norm": 0.07195625454187393,
|
| 1646 |
+
"learning_rate": 2.0608444788044744e-05,
|
| 1647 |
+
"loss": 0.02832261621952057,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.4694937044270193,
|
| 1652 |
+
"grad_norm": 0.05917816609144211,
|
| 1653 |
+
"learning_rate": 2.0423642106821118e-05,
|
| 1654 |
+
"loss": 0.017405569553375244,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.4757745566601568,
|
| 1659 |
+
"grad_norm": 0.050064459443092346,
|
| 1660 |
+
"learning_rate": 2.0239097804677755e-05,
|
| 1661 |
+
"loss": 0.03790850043296814,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.4820554088932942,
|
| 1666 |
+
"grad_norm": 0.06861293315887451,
|
| 1667 |
+
"learning_rate": 2.005482230090181e-05,
|
| 1668 |
+
"loss": 0.02318151295185089,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.4883362611264315,
|
| 1673 |
+
"grad_norm": 0.5692355036735535,
|
| 1674 |
+
"learning_rate": 1.987082599960423e-05,
|
| 1675 |
+
"loss": 0.05805472135543823,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.494617113359569,
|
| 1680 |
+
"grad_norm": 0.10575691610574722,
|
| 1681 |
+
"learning_rate": 1.9687119289132286e-05,
|
| 1682 |
+
"loss": 0.025386866927146912,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.5008979655927064,
|
| 1687 |
+
"grad_norm": 0.07561904937028885,
|
| 1688 |
+
"learning_rate": 1.9503712541483105e-05,
|
| 1689 |
+
"loss": 0.02556038498878479,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.5071788178258436,
|
| 1694 |
+
"grad_norm": 0.06342057138681412,
|
| 1695 |
+
"learning_rate": 1.9320616111718055e-05,
|
| 1696 |
+
"loss": 0.026987478137016296,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.513459670058981,
|
| 1701 |
+
"grad_norm": 0.22881099581718445,
|
| 1702 |
+
"learning_rate": 1.913784033737807e-05,
|
| 1703 |
+
"loss": 0.023304201662540436,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.5197405222921185,
|
| 1708 |
+
"grad_norm": 0.34815022349357605,
|
| 1709 |
+
"learning_rate": 1.8955395537900045e-05,
|
| 1710 |
+
"loss": 0.023628510534763336,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.5260213745252558,
|
| 1715 |
+
"grad_norm": 0.07004115730524063,
|
| 1716 |
+
"learning_rate": 1.877329201403417e-05,
|
| 1717 |
+
"loss": 0.01844399720430374,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.5323022267583932,
|
| 1722 |
+
"grad_norm": 0.5382964015007019,
|
| 1723 |
+
"learning_rate": 1.8591540047262408e-05,
|
| 1724 |
+
"loss": 0.022429858148097993,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.5385830789915307,
|
| 1729 |
+
"grad_norm": 0.10155873745679855,
|
| 1730 |
+
"learning_rate": 1.841014989921792e-05,
|
| 1731 |
+
"loss": 0.019565099477767946,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.544863931224668,
|
| 1736 |
+
"grad_norm": 0.45096078515052795,
|
| 1737 |
+
"learning_rate": 1.8229131811105773e-05,
|
| 1738 |
+
"loss": 0.02386317253112793,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.5511447834578054,
|
| 1743 |
+
"grad_norm": 1.6493111848831177,
|
| 1744 |
+
"learning_rate": 1.8048496003124692e-05,
|
| 1745 |
+
"loss": 0.022837825119495392,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.5574256356909428,
|
| 1750 |
+
"grad_norm": 0.2704576849937439,
|
| 1751 |
+
"learning_rate": 1.786825267389005e-05,
|
| 1752 |
+
"loss": 0.02158147543668747,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.56370648792408,
|
| 1757 |
+
"grad_norm": 0.14172030985355377,
|
| 1758 |
+
"learning_rate": 1.768841199985804e-05,
|
| 1759 |
+
"loss": 0.02343100756406784,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.5699873401572177,
|
| 1764 |
+
"grad_norm": 0.3839413523674011,
|
| 1765 |
+
"learning_rate": 1.7508984134751115e-05,
|
| 1766 |
+
"loss": 0.0211731418967247,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.576268192390355,
|
| 1771 |
+
"grad_norm": 0.17986074090003967,
|
| 1772 |
+
"learning_rate": 1.7329979208984757e-05,
|
| 1773 |
+
"loss": 0.050633370876312256,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.5825490446234922,
|
| 1778 |
+
"grad_norm": 0.11192335188388824,
|
| 1779 |
+
"learning_rate": 1.715140732909545e-05,
|
| 1780 |
+
"loss": 0.019000351428985596,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.58882989685663,
|
| 1785 |
+
"grad_norm": 0.06556563079357147,
|
| 1786 |
+
"learning_rate": 1.697327857717012e-05,
|
| 1787 |
+
"loss": 0.031407642364501956,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.5951107490897671,
|
| 1792 |
+
"grad_norm": 0.5898942351341248,
|
| 1793 |
+
"learning_rate": 1.67956030102769e-05,
|
| 1794 |
+
"loss": 0.03214563131332397,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.6013916013229044,
|
| 1799 |
+
"grad_norm": 0.11632583290338516,
|
| 1800 |
+
"learning_rate": 1.6618390659897297e-05,
|
| 1801 |
+
"loss": 0.12849061489105223,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.607672453556042,
|
| 1806 |
+
"grad_norm": 0.26994606852531433,
|
| 1807 |
+
"learning_rate": 1.64416515313598e-05,
|
| 1808 |
+
"loss": 0.01800369322299957,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.6139533057891793,
|
| 1813 |
+
"grad_norm": 0.3220551908016205,
|
| 1814 |
+
"learning_rate": 1.626539560327505e-05,
|
| 1815 |
+
"loss": 0.022022050619125367,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 1.6202341580223165,
|
| 1820 |
+
"grad_norm": 0.16517284512519836,
|
| 1821 |
+
"learning_rate": 1.6089632826972395e-05,
|
| 1822 |
+
"loss": 0.015372823178768157,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.6265150102554542,
|
| 1827 |
+
"grad_norm": 0.03851303830742836,
|
| 1828 |
+
"learning_rate": 1.591437312593804e-05,
|
| 1829 |
+
"loss": 0.0164533868432045,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.6327958624885914,
|
| 1834 |
+
"grad_norm": 0.3491153419017792,
|
| 1835 |
+
"learning_rate": 1.5739626395254798e-05,
|
| 1836 |
+
"loss": 0.031122979521751405,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 1.6390767147217287,
|
| 1841 |
+
"grad_norm": 0.06409186869859695,
|
| 1842 |
+
"learning_rate": 1.556540250104343e-05,
|
| 1843 |
+
"loss": 0.0328720360994339,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.6453575669548663,
|
| 1848 |
+
"grad_norm": 0.04858129099011421,
|
| 1849 |
+
"learning_rate": 1.539171127990557e-05,
|
| 1850 |
+
"loss": 0.05794671773910522,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.6516384191880036,
|
| 1855 |
+
"grad_norm": 0.262850821018219,
|
| 1856 |
+
"learning_rate": 1.5218562538368364e-05,
|
| 1857 |
+
"loss": 0.014299227297306061,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.6579192714211408,
|
| 1862 |
+
"grad_norm": 0.12737710773944855,
|
| 1863 |
+
"learning_rate": 1.504596605233084e-05,
|
| 1864 |
+
"loss": 0.02150038182735443,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 1.6642001236542785,
|
| 1869 |
+
"grad_norm": 0.06830194592475891,
|
| 1870 |
+
"learning_rate": 1.4873931566511901e-05,
|
| 1871 |
+
"loss": 0.01579444855451584,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.6704809758874157,
|
| 1876 |
+
"grad_norm": 0.08396535366773605,
|
| 1877 |
+
"learning_rate": 1.4702468793900188e-05,
|
| 1878 |
+
"loss": 0.027909812331199647,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 1.676761828120553,
|
| 1883 |
+
"grad_norm": 0.26403504610061646,
|
| 1884 |
+
"learning_rate": 1.4531587415205667e-05,
|
| 1885 |
+
"loss": 0.02287571579217911,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.6830426803536906,
|
| 1890 |
+
"grad_norm": 0.43412601947784424,
|
| 1891 |
+
"learning_rate": 1.436129707831308e-05,
|
| 1892 |
+
"loss": 0.017245300114154816,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.6893235325868279,
|
| 1897 |
+
"grad_norm": 0.18354500830173492,
|
| 1898 |
+
"learning_rate": 1.4191607397737206e-05,
|
| 1899 |
+
"loss": 0.031501269340515135,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.6956043848199651,
|
| 1904 |
+
"grad_norm": 0.04084882140159607,
|
| 1905 |
+
"learning_rate": 1.402252795408004e-05,
|
| 1906 |
+
"loss": 0.036852356791496274,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.7018852370531028,
|
| 1911 |
+
"grad_norm": 0.06115623563528061,
|
| 1912 |
+
"learning_rate": 1.3854068293489909e-05,
|
| 1913 |
+
"loss": 0.028507155179977418,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.70816608928624,
|
| 1918 |
+
"grad_norm": 0.32743072509765625,
|
| 1919 |
+
"learning_rate": 1.368623792712243e-05,
|
| 1920 |
+
"loss": 0.014020594954490661,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.7144469415193773,
|
| 1925 |
+
"grad_norm": 0.06222687289118767,
|
| 1926 |
+
"learning_rate": 1.3519046330603584e-05,
|
| 1927 |
+
"loss": 0.05104716420173645,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.720727793752515,
|
| 1932 |
+
"grad_norm": 0.06253579258918762,
|
| 1933 |
+
"learning_rate": 1.3352502943494672e-05,
|
| 1934 |
+
"loss": 0.026476505398750304,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.7270086459856522,
|
| 1939 |
+
"grad_norm": 0.4662589430809021,
|
| 1940 |
+
"learning_rate": 1.318661716875942e-05,
|
| 1941 |
+
"loss": 0.01999303698539734,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.7332894982187894,
|
| 1946 |
+
"grad_norm": 0.1247578039765358,
|
| 1947 |
+
"learning_rate": 1.3021398372233035e-05,
|
| 1948 |
+
"loss": 0.02069215327501297,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 1.739570350451927,
|
| 1953 |
+
"grad_norm": 0.3788617253303528,
|
| 1954 |
+
"learning_rate": 1.2856855882093432e-05,
|
| 1955 |
+
"loss": 0.018105532228946685,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.7458512026850643,
|
| 1960 |
+
"grad_norm": 1.8883055448532104,
|
| 1961 |
+
"learning_rate": 1.2692998988334612e-05,
|
| 1962 |
+
"loss": 0.04176854789257049,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 1.7521320549182016,
|
| 1967 |
+
"grad_norm": 0.21989472210407257,
|
| 1968 |
+
"learning_rate": 1.252983694224208e-05,
|
| 1969 |
+
"loss": 0.020176875591278075,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.7584129071513392,
|
| 1974 |
+
"grad_norm": 0.16766223311424255,
|
| 1975 |
+
"learning_rate": 1.236737895587056e-05,
|
| 1976 |
+
"loss": 0.020288291573524474,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.7646937593844765,
|
| 1981 |
+
"grad_norm": 0.09795085340738297,
|
| 1982 |
+
"learning_rate": 1.2205634201523922e-05,
|
| 1983 |
+
"loss": 0.015002159774303437,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.7709746116176137,
|
| 1988 |
+
"grad_norm": 0.029697278514504433,
|
| 1989 |
+
"learning_rate": 1.2044611811237242e-05,
|
| 1990 |
+
"loss": 0.023318004608154298,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.7772554638507514,
|
| 1995 |
+
"grad_norm": 0.3376557528972626,
|
| 1996 |
+
"learning_rate": 1.1884320876261268e-05,
|
| 1997 |
+
"loss": 0.0196805402636528,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.7835363160838886,
|
| 2002 |
+
"grad_norm": 0.06531202793121338,
|
| 2003 |
+
"learning_rate": 1.17247704465491e-05,
|
| 2004 |
+
"loss": 0.020646215975284578,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.7898171683170259,
|
| 2009 |
+
"grad_norm": 0.36572521924972534,
|
| 2010 |
+
"learning_rate": 1.1565969530245277e-05,
|
| 2011 |
+
"loss": 0.025085961818695067,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.7960980205501635,
|
| 2016 |
+
"grad_norm": 0.1814069002866745,
|
| 2017 |
+
"learning_rate": 1.140792709317713e-05,
|
| 2018 |
+
"loss": 0.030590158700942994,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.8023788727833008,
|
| 2023 |
+
"grad_norm": 0.24981825053691864,
|
| 2024 |
+
"learning_rate": 1.12506520583486e-05,
|
| 2025 |
+
"loss": 0.0215068444609642,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 1.808659725016438,
|
| 2030 |
+
"grad_norm": 0.35276567935943604,
|
| 2031 |
+
"learning_rate": 1.1094153305436453e-05,
|
| 2032 |
+
"loss": 0.01926688849925995,
|
| 2033 |
+
"step": 5760
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 1.8149405772495757,
|
| 2037 |
+
"grad_norm": 0.02407790906727314,
|
| 2038 |
+
"learning_rate": 1.0938439670288928e-05,
|
| 2039 |
+
"loss": 0.013058166205883025,
|
| 2040 |
+
"step": 5780
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.821221429482713,
|
| 2044 |
+
"grad_norm": 0.08507027477025986,
|
| 2045 |
+
"learning_rate": 1.0783519944426874e-05,
|
| 2046 |
+
"loss": 0.01877378821372986,
|
| 2047 |
+
"step": 5800
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.8275022817158502,
|
| 2051 |
+
"grad_norm": 0.18674074113368988,
|
| 2052 |
+
"learning_rate": 1.0629402874547378e-05,
|
| 2053 |
+
"loss": 0.018952280282974243,
|
| 2054 |
+
"step": 5820
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.8337831339489878,
|
| 2058 |
+
"grad_norm": 0.057133324444293976,
|
| 2059 |
+
"learning_rate": 1.047609716202996e-05,
|
| 2060 |
+
"loss": 0.030013075470924376,
|
| 2061 |
+
"step": 5840
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.840063986182125,
|
| 2065 |
+
"grad_norm": 0.10284405946731567,
|
| 2066 |
+
"learning_rate": 1.0323611462445246e-05,
|
| 2067 |
+
"loss": 0.020608581602573395,
|
| 2068 |
+
"step": 5860
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 1.8463448384152623,
|
| 2072 |
+
"grad_norm": 0.25261375308036804,
|
| 2073 |
+
"learning_rate": 1.0171954385066303e-05,
|
| 2074 |
+
"loss": 0.02777992784976959,
|
| 2075 |
+
"step": 5880
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 1.8526256906484,
|
| 2079 |
+
"grad_norm": 0.38947853446006775,
|
| 2080 |
+
"learning_rate": 1.0021134492382606e-05,
|
| 2081 |
+
"loss": 0.02200171649456024,
|
| 2082 |
+
"step": 5900
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 1.8589065428815372,
|
| 2086 |
+
"grad_norm": 0.1173999160528183,
|
| 2087 |
+
"learning_rate": 9.87116029961652e-06,
|
| 2088 |
+
"loss": 0.019530104100704195,
|
| 2089 |
+
"step": 5920
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 1.8651873951146745,
|
| 2093 |
+
"grad_norm": 2.3180384635925293,
|
| 2094 |
+
"learning_rate": 9.722040274242594e-06,
|
| 2095 |
+
"loss": 0.028142544627189636,
|
| 2096 |
+
"step": 5940
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 1.8714682473478121,
|
| 2100 |
+
"grad_norm": 0.4677383601665497,
|
| 2101 |
+
"learning_rate": 9.573782835509471e-06,
|
| 2102 |
+
"loss": 0.017200836539268495,
|
| 2103 |
+
"step": 5960
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.8777490995809494,
|
| 2107 |
+
"grad_norm": 0.19456204771995544,
|
| 2108 |
+
"learning_rate": 9.426396353964565e-06,
|
| 2109 |
+
"loss": 0.015525393187999725,
|
| 2110 |
+
"step": 5980
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 1.8840299518140866,
|
| 2114 |
+
"grad_norm": 0.561476469039917,
|
| 2115 |
+
"learning_rate": 9.279889150981421e-06,
|
| 2116 |
+
"loss": 0.04386698603630066,
|
| 2117 |
+
"step": 6000
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 1.8903108040472243,
|
| 2121 |
+
"grad_norm": 0.0613991916179657,
|
| 2122 |
+
"learning_rate": 9.134269498289922e-06,
|
| 2123 |
+
"loss": 0.031869342923164366,
|
| 2124 |
+
"step": 6020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 1.8965916562803615,
|
| 2128 |
+
"grad_norm": 0.047774072736501694,
|
| 2129 |
+
"learning_rate": 8.989545617509283e-06,
|
| 2130 |
+
"loss": 0.02780761420726776,
|
| 2131 |
+
"step": 6040
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 1.9028725085134988,
|
| 2135 |
+
"grad_norm": 0.33418071269989014,
|
| 2136 |
+
"learning_rate": 8.845725679683822e-06,
|
| 2137 |
+
"loss": 0.04635312259197235,
|
| 2138 |
+
"step": 6060
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 1.9091533607466364,
|
| 2142 |
+
"grad_norm": 0.0721767470240593,
|
| 2143 |
+
"learning_rate": 8.702817804821664e-06,
|
| 2144 |
+
"loss": 0.01725383847951889,
|
| 2145 |
+
"step": 6080
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.9154342129797737,
|
| 2149 |
+
"grad_norm": 0.04245443269610405,
|
| 2150 |
+
"learning_rate": 8.560830061436265e-06,
|
| 2151 |
+
"loss": 0.02144094407558441,
|
| 2152 |
+
"step": 6100
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 1.921715065212911,
|
| 2156 |
+
"grad_norm": 0.1140967458486557,
|
| 2157 |
+
"learning_rate": 8.419770466090896e-06,
|
| 2158 |
+
"loss": 0.014450745284557342,
|
| 2159 |
+
"step": 6120
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 1.9279959174460486,
|
| 2163 |
+
"grad_norm": 0.06213388964533806,
|
| 2164 |
+
"learning_rate": 8.27964698294599e-06,
|
| 2165 |
+
"loss": 0.023716664314270018,
|
| 2166 |
+
"step": 6140
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 1.9342767696791858,
|
| 2170 |
+
"grad_norm": 0.2430906891822815,
|
| 2171 |
+
"learning_rate": 8.140467523309527e-06,
|
| 2172 |
+
"loss": 0.011777687072753906,
|
| 2173 |
+
"step": 6160
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 1.9405576219123233,
|
| 2177 |
+
"grad_norm": 0.044944941997528076,
|
| 2178 |
+
"learning_rate": 8.002239945190345e-06,
|
| 2179 |
+
"loss": 0.01790996491909027,
|
| 2180 |
+
"step": 6180
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 1.9468384741454607,
|
| 2184 |
+
"grad_norm": 0.032080262899398804,
|
| 2185 |
+
"learning_rate": 7.864972052854485e-06,
|
| 2186 |
+
"loss": 0.03710399568080902,
|
| 2187 |
+
"step": 6200
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.953119326378598,
|
| 2191 |
+
"grad_norm": 0.4400108754634857,
|
| 2192 |
+
"learning_rate": 7.72867159638456e-06,
|
| 2193 |
+
"loss": 0.022929078340530394,
|
| 2194 |
+
"step": 6220
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 1.9594001786117354,
|
| 2198 |
+
"grad_norm": 0.04446094483137131,
|
| 2199 |
+
"learning_rate": 7.593346271242216e-06,
|
| 2200 |
+
"loss": 0.01109272986650467,
|
| 2201 |
+
"step": 6240
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 1.965681030844873,
|
| 2205 |
+
"grad_norm": 0.37326884269714355,
|
| 2206 |
+
"learning_rate": 7.459003717833607e-06,
|
| 2207 |
+
"loss": 0.04479179680347443,
|
| 2208 |
+
"step": 6260
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 1.9719618830780101,
|
| 2212 |
+
"grad_norm": 0.03807030990719795,
|
| 2213 |
+
"learning_rate": 7.325651521078053e-06,
|
| 2214 |
+
"loss": 0.020693680644035338,
|
| 2215 |
+
"step": 6280
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 1.9782427353111476,
|
| 2219 |
+
"grad_norm": 0.07757691293954849,
|
| 2220 |
+
"learning_rate": 7.193297209979777e-06,
|
| 2221 |
+
"loss": 0.015228751301765441,
|
| 2222 |
+
"step": 6300
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 1.984523587544285,
|
| 2226 |
+
"grad_norm": 0.06824570149183273,
|
| 2227 |
+
"learning_rate": 7.061948257202858e-06,
|
| 2228 |
+
"loss": 0.01840972900390625,
|
| 2229 |
+
"step": 6320
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 1.9908044397774223,
|
| 2233 |
+
"grad_norm": 0.1820923089981079,
|
| 2234 |
+
"learning_rate": 6.931612078649283e-06,
|
| 2235 |
+
"loss": 0.02284184694290161,
|
| 2236 |
+
"step": 6340
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 1.9970852920105597,
|
| 2240 |
+
"grad_norm": 0.0430208295583725,
|
| 2241 |
+
"learning_rate": 6.8022960330402616e-06,
|
| 2242 |
+
"loss": 0.016644990444183348,
|
| 2243 |
+
"step": 6360
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.003140426116569,
|
| 2247 |
+
"grad_norm": 0.050990670919418335,
|
| 2248 |
+
"learning_rate": 6.674007421500786e-06,
|
| 2249 |
+
"loss": 0.03776961266994476,
|
| 2250 |
+
"step": 6380
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.009421278349706,
|
| 2254 |
+
"grad_norm": 0.13200590014457703,
|
| 2255 |
+
"learning_rate": 6.546753487147375e-06,
|
| 2256 |
+
"loss": 0.026686957478523253,
|
| 2257 |
+
"step": 6400
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.0157021305828433,
|
| 2261 |
+
"grad_norm": 0.08269485086202621,
|
| 2262 |
+
"learning_rate": 6.42054141467914e-06,
|
| 2263 |
+
"loss": 0.01832299828529358,
|
| 2264 |
+
"step": 6420
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.021982982815981,
|
| 2268 |
+
"grad_norm": 0.04629537835717201,
|
| 2269 |
+
"learning_rate": 6.2953783299721486e-06,
|
| 2270 |
+
"loss": 0.01113806888461113,
|
| 2271 |
+
"step": 6440
|
| 2272 |
+
}
|
| 2273 |
+
],
|
| 2274 |
+
"logging_steps": 20,
|
| 2275 |
+
"max_steps": 8362,
|
| 2276 |
+
"num_input_tokens_seen": 0,
|
| 2277 |
+
"num_train_epochs": 3,
|
| 2278 |
+
"save_steps": 1000000000,
|
| 2279 |
+
"stateful_callbacks": {
|
| 2280 |
+
"TrainerControl": {
|
| 2281 |
+
"args": {
|
| 2282 |
+
"should_epoch_stop": false,
|
| 2283 |
+
"should_evaluate": false,
|
| 2284 |
+
"should_log": false,
|
| 2285 |
+
"should_save": true,
|
| 2286 |
+
"should_training_stop": false
|
| 2287 |
+
},
|
| 2288 |
+
"attributes": {}
|
| 2289 |
+
}
|
| 2290 |
+
},
|
| 2291 |
+
"total_flos": 2.038940224187904e+16,
|
| 2292 |
+
"train_batch_size": 1,
|
| 2293 |
+
"trial_name": null,
|
| 2294 |
+
"trial_params": null
|
| 2295 |
+
}
|