Upload folder using huggingface_hub
Browse files
slots/9/checkpoint-5809/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/9/checkpoint-5809/trainer_state.json
ADDED
|
@@ -0,0 +1,2071 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.431702497939102,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 5809,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.00041872211245305734,
|
| 14 |
+
"grad_norm": 1.9653303623199463,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 0.807927668094635,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.008374442249061146,
|
| 21 |
+
"grad_norm": 1.5293045043945312,
|
| 22 |
+
"learning_rate": 4.999942834660501e-05,
|
| 23 |
+
"loss": 1.1682350761012028,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.016748884498122292,
|
| 28 |
+
"grad_norm": 2.27285099029541,
|
| 29 |
+
"learning_rate": 4.999745229435885e-05,
|
| 30 |
+
"loss": 0.8465386390686035,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.025123326747183438,
|
| 35 |
+
"grad_norm": 2.492253541946411,
|
| 36 |
+
"learning_rate": 4.999406489735261e-05,
|
| 37 |
+
"loss": 1.2165610313415527,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.033497768996244584,
|
| 42 |
+
"grad_norm": 2.0843803882598877,
|
| 43 |
+
"learning_rate": 4.99892663468372e-05,
|
| 44 |
+
"loss": 0.8231470108032226,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.041872211245305734,
|
| 49 |
+
"grad_norm": 2.5146677494049072,
|
| 50 |
+
"learning_rate": 4.998340088876368e-05,
|
| 51 |
+
"loss": 0.8773845672607422,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.050246653494366876,
|
| 56 |
+
"grad_norm": 2.7516226768493652,
|
| 57 |
+
"learning_rate": 4.997585144070399e-05,
|
| 58 |
+
"loss": 0.8910114288330078,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.058621095743428026,
|
| 63 |
+
"grad_norm": 10.78797435760498,
|
| 64 |
+
"learning_rate": 4.996689186745859e-05,
|
| 65 |
+
"loss": 0.6811366558074952,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.06699553799248917,
|
| 70 |
+
"grad_norm": 2.3442978858947754,
|
| 71 |
+
"learning_rate": 4.995707460425363e-05,
|
| 72 |
+
"loss": 0.8841165542602539,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.07536998024155032,
|
| 77 |
+
"grad_norm": 1.5972583293914795,
|
| 78 |
+
"learning_rate": 4.994598565925399e-05,
|
| 79 |
+
"loss": 1.0401561737060547,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.08374442249061147,
|
| 84 |
+
"grad_norm": 1.4799625873565674,
|
| 85 |
+
"learning_rate": 4.9932939868583476e-05,
|
| 86 |
+
"loss": 0.6746731758117676,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09211886473967261,
|
| 91 |
+
"grad_norm": 17.09868812561035,
|
| 92 |
+
"learning_rate": 4.991848637549505e-05,
|
| 93 |
+
"loss": 0.7095736503601074,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.10049330698873375,
|
| 98 |
+
"grad_norm": 9.661768913269043,
|
| 99 |
+
"learning_rate": 4.990345241504926e-05,
|
| 100 |
+
"loss": 0.7918826580047608,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.10886774923779491,
|
| 105 |
+
"grad_norm": 2.0028388500213623,
|
| 106 |
+
"learning_rate": 4.98862563217225e-05,
|
| 107 |
+
"loss": 0.6271281242370605,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.11724219148685605,
|
| 112 |
+
"grad_norm": 27.938236236572266,
|
| 113 |
+
"learning_rate": 4.9867655161709595e-05,
|
| 114 |
+
"loss": 0.8460742950439453,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1256166337359172,
|
| 119 |
+
"grad_norm": 2.7466659545898438,
|
| 120 |
+
"learning_rate": 4.984868357232379e-05,
|
| 121 |
+
"loss": 0.7270052433013916,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.13399107598497834,
|
| 126 |
+
"grad_norm": 0.8086409568786621,
|
| 127 |
+
"learning_rate": 4.982734562514903e-05,
|
| 128 |
+
"loss": 0.6195536136627198,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.14236551823403948,
|
| 133 |
+
"grad_norm": 1.1716150045394897,
|
| 134 |
+
"learning_rate": 4.98046059373592e-05,
|
| 135 |
+
"loss": 0.6283158779144287,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.15073996048310065,
|
| 140 |
+
"grad_norm": 6.236642837524414,
|
| 141 |
+
"learning_rate": 4.97804657928267e-05,
|
| 142 |
+
"loss": 0.5698219299316406,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1591144027321618,
|
| 147 |
+
"grad_norm": 1.4684100151062012,
|
| 148 |
+
"learning_rate": 4.9754926554493095e-05,
|
| 149 |
+
"loss": 0.5527248382568359,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.16748884498122293,
|
| 154 |
+
"grad_norm": 0.8660781979560852,
|
| 155 |
+
"learning_rate": 4.972798966429213e-05,
|
| 156 |
+
"loss": 0.48558568954467773,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.17586328723028408,
|
| 161 |
+
"grad_norm": 0.9932934641838074,
|
| 162 |
+
"learning_rate": 4.969965664306835e-05,
|
| 163 |
+
"loss": 0.46672744750976564,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.18423772947934522,
|
| 168 |
+
"grad_norm": 1.0187177658081055,
|
| 169 |
+
"learning_rate": 4.966992909049121e-05,
|
| 170 |
+
"loss": 0.4634897232055664,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.19261217172840636,
|
| 175 |
+
"grad_norm": 1.1465648412704468,
|
| 176 |
+
"learning_rate": 4.9638808684964765e-05,
|
| 177 |
+
"loss": 0.4585463523864746,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.2009866139774675,
|
| 182 |
+
"grad_norm": 1.9938956499099731,
|
| 183 |
+
"learning_rate": 4.960629718353295e-05,
|
| 184 |
+
"loss": 0.5669815540313721,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.20936105622652867,
|
| 189 |
+
"grad_norm": 2.2749507427215576,
|
| 190 |
+
"learning_rate": 4.9572396421780296e-05,
|
| 191 |
+
"loss": 0.4630398750305176,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.21773549847558982,
|
| 196 |
+
"grad_norm": 1.0254639387130737,
|
| 197 |
+
"learning_rate": 4.9537108313728365e-05,
|
| 198 |
+
"loss": 0.409474515914917,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.22610994072465096,
|
| 203 |
+
"grad_norm": 1.2044686079025269,
|
| 204 |
+
"learning_rate": 4.9500434851727635e-05,
|
| 205 |
+
"loss": 0.4084805965423584,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.2344843829737121,
|
| 210 |
+
"grad_norm": 1.0856112241744995,
|
| 211 |
+
"learning_rate": 4.9462378106345064e-05,
|
| 212 |
+
"loss": 0.3315908908843994,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.24285882522277324,
|
| 217 |
+
"grad_norm": 0.9389876127243042,
|
| 218 |
+
"learning_rate": 4.9422940226247135e-05,
|
| 219 |
+
"loss": 0.4294145584106445,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.2512332674718344,
|
| 224 |
+
"grad_norm": 1.2819859981536865,
|
| 225 |
+
"learning_rate": 4.9382123438078595e-05,
|
| 226 |
+
"loss": 0.374059796333313,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.25960770972089553,
|
| 231 |
+
"grad_norm": 1.4377272129058838,
|
| 232 |
+
"learning_rate": 4.933993004633669e-05,
|
| 233 |
+
"loss": 0.3626229286193848,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.2679821519699567,
|
| 238 |
+
"grad_norm": 2.0702695846557617,
|
| 239 |
+
"learning_rate": 4.929636243324107e-05,
|
| 240 |
+
"loss": 0.5260854244232178,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.2763565942190178,
|
| 245 |
+
"grad_norm": 1.4993720054626465,
|
| 246 |
+
"learning_rate": 4.925142305859931e-05,
|
| 247 |
+
"loss": 0.41648120880126954,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.28473103646807896,
|
| 252 |
+
"grad_norm": 0.9975736737251282,
|
| 253 |
+
"learning_rate": 4.920511445966803e-05,
|
| 254 |
+
"loss": 0.45428853034973143,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.29310547871714016,
|
| 259 |
+
"grad_norm": 1.5575836896896362,
|
| 260 |
+
"learning_rate": 4.91574392510096e-05,
|
| 261 |
+
"loss": 0.36581900119781496,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.3014799209662013,
|
| 266 |
+
"grad_norm": 0.9830530881881714,
|
| 267 |
+
"learning_rate": 4.9108400124344564e-05,
|
| 268 |
+
"loss": 0.3410146951675415,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.30985436321526244,
|
| 273 |
+
"grad_norm": 1.0295761823654175,
|
| 274 |
+
"learning_rate": 4.905799984839966e-05,
|
| 275 |
+
"loss": 0.5368211269378662,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3182288054643236,
|
| 280 |
+
"grad_norm": 0.9159882068634033,
|
| 281 |
+
"learning_rate": 4.900624126875149e-05,
|
| 282 |
+
"loss": 0.37833280563354493,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3266032477133847,
|
| 287 |
+
"grad_norm": 0.5635650157928467,
|
| 288 |
+
"learning_rate": 4.895312730766587e-05,
|
| 289 |
+
"loss": 0.4511431694030762,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.33497768996244587,
|
| 294 |
+
"grad_norm": 1.7489354610443115,
|
| 295 |
+
"learning_rate": 4.8898660963932815e-05,
|
| 296 |
+
"loss": 0.32177019119262695,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.343352132211507,
|
| 301 |
+
"grad_norm": 9.12048053741455,
|
| 302 |
+
"learning_rate": 4.884284531269727e-05,
|
| 303 |
+
"loss": 0.32772254943847656,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.35172657446056815,
|
| 308 |
+
"grad_norm": 1.3818485736846924,
|
| 309 |
+
"learning_rate": 4.878568350528546e-05,
|
| 310 |
+
"loss": 0.4475229740142822,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3601010167096293,
|
| 315 |
+
"grad_norm": 0.775784432888031,
|
| 316 |
+
"learning_rate": 4.872717876902696e-05,
|
| 317 |
+
"loss": 0.3400275230407715,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.36847545895869044,
|
| 322 |
+
"grad_norm": 1.5080106258392334,
|
| 323 |
+
"learning_rate": 4.866733440707252e-05,
|
| 324 |
+
"loss": 0.2904891729354858,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3768499012077516,
|
| 329 |
+
"grad_norm": 0.9386821389198303,
|
| 330 |
+
"learning_rate": 4.860615379820752e-05,
|
| 331 |
+
"loss": 0.3725273609161377,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3852243434568127,
|
| 336 |
+
"grad_norm": 1.6192686557769775,
|
| 337 |
+
"learning_rate": 4.854364039666125e-05,
|
| 338 |
+
"loss": 0.5561938762664795,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.39359878570587387,
|
| 343 |
+
"grad_norm": 0.4697323739528656,
|
| 344 |
+
"learning_rate": 4.847979773191186e-05,
|
| 345 |
+
"loss": 0.37938408851623534,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.401973227954935,
|
| 350 |
+
"grad_norm": 0.9349277019500732,
|
| 351 |
+
"learning_rate": 4.84146294084871e-05,
|
| 352 |
+
"loss": 0.27554936408996583,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.4103476702039962,
|
| 357 |
+
"grad_norm": 1.1451354026794434,
|
| 358 |
+
"learning_rate": 4.834813910576079e-05,
|
| 359 |
+
"loss": 0.4046372413635254,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.41872211245305735,
|
| 364 |
+
"grad_norm": 1.3403191566467285,
|
| 365 |
+
"learning_rate": 4.828033057774512e-05,
|
| 366 |
+
"loss": 0.36427290439605714,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.4270965547021185,
|
| 371 |
+
"grad_norm": 0.5559656620025635,
|
| 372 |
+
"learning_rate": 4.821120765287866e-05,
|
| 373 |
+
"loss": 0.2821886301040649,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.43547099695117963,
|
| 378 |
+
"grad_norm": 3.9479188919067383,
|
| 379 |
+
"learning_rate": 4.814077423381024e-05,
|
| 380 |
+
"loss": 0.31128578186035155,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.4438454392002408,
|
| 385 |
+
"grad_norm": 1.6531774997711182,
|
| 386 |
+
"learning_rate": 4.806903429717857e-05,
|
| 387 |
+
"loss": 0.29161553382873534,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.4522198814493019,
|
| 392 |
+
"grad_norm": 0.7019716501235962,
|
| 393 |
+
"learning_rate": 4.7995991893387796e-05,
|
| 394 |
+
"loss": 0.25183801651000975,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.46059432369836306,
|
| 399 |
+
"grad_norm": 2.483739137649536,
|
| 400 |
+
"learning_rate": 4.792165114637872e-05,
|
| 401 |
+
"loss": 0.2702826738357544,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.4689687659474242,
|
| 406 |
+
"grad_norm": 1.0199668407440186,
|
| 407 |
+
"learning_rate": 4.784601625339602e-05,
|
| 408 |
+
"loss": 0.2821660041809082,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.47734320819648535,
|
| 413 |
+
"grad_norm": 1.9553228616714478,
|
| 414 |
+
"learning_rate": 4.776909148475128e-05,
|
| 415 |
+
"loss": 0.24843573570251465,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.4857176504455465,
|
| 420 |
+
"grad_norm": 0.8806698322296143,
|
| 421 |
+
"learning_rate": 4.769088118358187e-05,
|
| 422 |
+
"loss": 0.20631277561187744,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.49409209269460763,
|
| 427 |
+
"grad_norm": 0.3777574896812439,
|
| 428 |
+
"learning_rate": 4.761138976560574e-05,
|
| 429 |
+
"loss": 0.2833678960800171,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5024665349436688,
|
| 434 |
+
"grad_norm": 0.6823768615722656,
|
| 435 |
+
"learning_rate": 4.7530621718872084e-05,
|
| 436 |
+
"loss": 0.3575673818588257,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5108409771927299,
|
| 441 |
+
"grad_norm": 0.5605328679084778,
|
| 442 |
+
"learning_rate": 4.744858160350803e-05,
|
| 443 |
+
"loss": 0.29072275161743166,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5192154194417911,
|
| 448 |
+
"grad_norm": 0.6820014715194702,
|
| 449 |
+
"learning_rate": 4.736527405146107e-05,
|
| 450 |
+
"loss": 0.2855283498764038,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5275898616908522,
|
| 455 |
+
"grad_norm": 0.5537764430046082,
|
| 456 |
+
"learning_rate": 4.728070376623761e-05,
|
| 457 |
+
"loss": 0.24044954776763916,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5359643039399133,
|
| 462 |
+
"grad_norm": 0.6336297988891602,
|
| 463 |
+
"learning_rate": 4.719487552263741e-05,
|
| 464 |
+
"loss": 0.24357831478118896,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5443387461889745,
|
| 469 |
+
"grad_norm": 0.8495275974273682,
|
| 470 |
+
"learning_rate": 4.710779416648397e-05,
|
| 471 |
+
"loss": 0.2741832971572876,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.5527131884380356,
|
| 476 |
+
"grad_norm": 0.9994595646858215,
|
| 477 |
+
"learning_rate": 4.701946461435095e-05,
|
| 478 |
+
"loss": 0.2851997375488281,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.5610876306870968,
|
| 483 |
+
"grad_norm": 1.2617225646972656,
|
| 484 |
+
"learning_rate": 4.69298918532846e-05,
|
| 485 |
+
"loss": 0.30367312431335447,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.5694620729361579,
|
| 490 |
+
"grad_norm": 1.1010406017303467,
|
| 491 |
+
"learning_rate": 4.6839080940522185e-05,
|
| 492 |
+
"loss": 0.26403853893280027,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.5778365151852191,
|
| 497 |
+
"grad_norm": 1.0911697149276733,
|
| 498 |
+
"learning_rate": 4.674703700320641e-05,
|
| 499 |
+
"loss": 0.2289508104324341,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5862109574342803,
|
| 504 |
+
"grad_norm": 0.9599721431732178,
|
| 505 |
+
"learning_rate": 4.6653765238096037e-05,
|
| 506 |
+
"loss": 0.20489723682403566,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5945853996833415,
|
| 511 |
+
"grad_norm": 1.2036864757537842,
|
| 512 |
+
"learning_rate": 4.655927091127239e-05,
|
| 513 |
+
"loss": 0.2418372631072998,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6029598419324026,
|
| 518 |
+
"grad_norm": 0.828085720539093,
|
| 519 |
+
"learning_rate": 4.646355935784208e-05,
|
| 520 |
+
"loss": 0.20368292331695556,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6113342841814637,
|
| 525 |
+
"grad_norm": 0.7752611637115479,
|
| 526 |
+
"learning_rate": 4.6366635981635786e-05,
|
| 527 |
+
"loss": 0.2782510042190552,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6197087264305249,
|
| 532 |
+
"grad_norm": 1.2014942169189453,
|
| 533 |
+
"learning_rate": 4.626850625490313e-05,
|
| 534 |
+
"loss": 0.3075052261352539,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.628083168679586,
|
| 539 |
+
"grad_norm": 0.5737164616584778,
|
| 540 |
+
"learning_rate": 4.6169175718003764e-05,
|
| 541 |
+
"loss": 0.17472189664840698,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6364576109286472,
|
| 546 |
+
"grad_norm": 0.6982183456420898,
|
| 547 |
+
"learning_rate": 4.60686499790945e-05,
|
| 548 |
+
"loss": 0.2059044361114502,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.6448320531777083,
|
| 553 |
+
"grad_norm": 0.3611967861652374,
|
| 554 |
+
"learning_rate": 4.596693471381272e-05,
|
| 555 |
+
"loss": 0.19336872100830077,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6532064954267695,
|
| 560 |
+
"grad_norm": 0.8547729849815369,
|
| 561 |
+
"learning_rate": 4.5864035664955954e-05,
|
| 562 |
+
"loss": 0.18282541036605834,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.6615809376758306,
|
| 567 |
+
"grad_norm": 0.3589843809604645,
|
| 568 |
+
"learning_rate": 4.5759958642157565e-05,
|
| 569 |
+
"loss": 0.22185680866241456,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.6699553799248917,
|
| 574 |
+
"grad_norm": 1.1286545991897583,
|
| 575 |
+
"learning_rate": 4.5654709521558816e-05,
|
| 576 |
+
"loss": 0.2895075559616089,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.6783298221739529,
|
| 581 |
+
"grad_norm": 0.7348793745040894,
|
| 582 |
+
"learning_rate": 4.554829424547708e-05,
|
| 583 |
+
"loss": 0.17382105588912963,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.686704264423014,
|
| 588 |
+
"grad_norm": 2.081247568130493,
|
| 589 |
+
"learning_rate": 4.544071882207033e-05,
|
| 590 |
+
"loss": 0.19009811878204347,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.6950787066720752,
|
| 595 |
+
"grad_norm": 0.14719480276107788,
|
| 596 |
+
"learning_rate": 4.533198932499791e-05,
|
| 597 |
+
"loss": 0.1196211576461792,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7034531489211363,
|
| 602 |
+
"grad_norm": 1.2243156433105469,
|
| 603 |
+
"learning_rate": 4.522211189307768e-05,
|
| 604 |
+
"loss": 0.25608577728271487,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.7118275911701974,
|
| 609 |
+
"grad_norm": 0.25937438011169434,
|
| 610 |
+
"learning_rate": 4.511109272993933e-05,
|
| 611 |
+
"loss": 0.19631800651550294,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7202020334192586,
|
| 616 |
+
"grad_norm": 2.1448802947998047,
|
| 617 |
+
"learning_rate": 4.4998938103674215e-05,
|
| 618 |
+
"loss": 0.2997403621673584,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7285764756683197,
|
| 623 |
+
"grad_norm": 0.7737797498703003,
|
| 624 |
+
"learning_rate": 4.4885654346481386e-05,
|
| 625 |
+
"loss": 0.17992446422576905,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7369509179173809,
|
| 630 |
+
"grad_norm": 3.7802436351776123,
|
| 631 |
+
"learning_rate": 4.477124785431014e-05,
|
| 632 |
+
"loss": 0.15774155855178834,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.745325360166442,
|
| 637 |
+
"grad_norm": 1.911975622177124,
|
| 638 |
+
"learning_rate": 4.465572508649886e-05,
|
| 639 |
+
"loss": 0.22157161235809325,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7536998024155032,
|
| 644 |
+
"grad_norm": 1.0170443058013916,
|
| 645 |
+
"learning_rate": 4.453909256541034e-05,
|
| 646 |
+
"loss": 0.16437970399856566,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.7620742446645643,
|
| 651 |
+
"grad_norm": 0.15334242582321167,
|
| 652 |
+
"learning_rate": 4.4421356876063566e-05,
|
| 653 |
+
"loss": 0.17869235277175904,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.7704486869136254,
|
| 658 |
+
"grad_norm": 0.8450778722763062,
|
| 659 |
+
"learning_rate": 4.430252466576188e-05,
|
| 660 |
+
"loss": 0.13349066972732543,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.7788231291626866,
|
| 665 |
+
"grad_norm": 0.18455669283866882,
|
| 666 |
+
"learning_rate": 4.418260264371768e-05,
|
| 667 |
+
"loss": 0.14980933666229249,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.7871975714117477,
|
| 672 |
+
"grad_norm": 2.0214483737945557,
|
| 673 |
+
"learning_rate": 4.4061597580673676e-05,
|
| 674 |
+
"loss": 0.2396923303604126,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.7955720136608089,
|
| 679 |
+
"grad_norm": 0.37638697028160095,
|
| 680 |
+
"learning_rate": 4.3939516308520555e-05,
|
| 681 |
+
"loss": 0.15193476676940917,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.80394645590987,
|
| 686 |
+
"grad_norm": 0.89265376329422,
|
| 687 |
+
"learning_rate": 4.3816365719911286e-05,
|
| 688 |
+
"loss": 0.15367662906646729,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8123208981589312,
|
| 693 |
+
"grad_norm": 1.2247042655944824,
|
| 694 |
+
"learning_rate": 4.3692152767871955e-05,
|
| 695 |
+
"loss": 0.12489446401596069,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8206953404079924,
|
| 700 |
+
"grad_norm": 0.18706317245960236,
|
| 701 |
+
"learning_rate": 4.3566884465409206e-05,
|
| 702 |
+
"loss": 0.21929187774658204,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.8290697826570536,
|
| 707 |
+
"grad_norm": 0.33726993203163147,
|
| 708 |
+
"learning_rate": 4.344056788511428e-05,
|
| 709 |
+
"loss": 0.17386707067489623,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.8374442249061147,
|
| 714 |
+
"grad_norm": 0.12756434082984924,
|
| 715 |
+
"learning_rate": 4.331321015876373e-05,
|
| 716 |
+
"loss": 0.11789515018463134,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.8458186671551758,
|
| 721 |
+
"grad_norm": 0.934647262096405,
|
| 722 |
+
"learning_rate": 4.31848184769167e-05,
|
| 723 |
+
"loss": 0.1382400155067444,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.854193109404237,
|
| 728 |
+
"grad_norm": 0.5541802644729614,
|
| 729 |
+
"learning_rate": 4.305540008850902e-05,
|
| 730 |
+
"loss": 0.34618980884552003,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.8625675516532981,
|
| 735 |
+
"grad_norm": 0.6069034934043884,
|
| 736 |
+
"learning_rate": 4.2924962300443904e-05,
|
| 737 |
+
"loss": 0.2154491662979126,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.8709419939023593,
|
| 742 |
+
"grad_norm": 0.3920958936214447,
|
| 743 |
+
"learning_rate": 4.2793512477179395e-05,
|
| 744 |
+
"loss": 0.1823054313659668,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.8793164361514204,
|
| 749 |
+
"grad_norm": 1.0203856229782104,
|
| 750 |
+
"learning_rate": 4.2661058040312584e-05,
|
| 751 |
+
"loss": 0.16364898681640624,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.8876908784004816,
|
| 756 |
+
"grad_norm": 0.2837757170200348,
|
| 757 |
+
"learning_rate": 4.2527606468160584e-05,
|
| 758 |
+
"loss": 0.16561037302017212,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.8960653206495427,
|
| 763 |
+
"grad_norm": 0.588916003704071,
|
| 764 |
+
"learning_rate": 4.2393165295338336e-05,
|
| 765 |
+
"loss": 0.11391867399215698,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.9044397628986038,
|
| 770 |
+
"grad_norm": 0.39494433999061584,
|
| 771 |
+
"learning_rate": 4.2257742112333145e-05,
|
| 772 |
+
"loss": 0.14886506795883178,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.912814205147665,
|
| 777 |
+
"grad_norm": 0.43156421184539795,
|
| 778 |
+
"learning_rate": 4.2121344565076204e-05,
|
| 779 |
+
"loss": 0.18619248867034913,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.9211886473967261,
|
| 784 |
+
"grad_norm": 0.643484890460968,
|
| 785 |
+
"learning_rate": 4.198398035451082e-05,
|
| 786 |
+
"loss": 0.18319352865219116,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.9295630896457873,
|
| 791 |
+
"grad_norm": 0.7975239157676697,
|
| 792 |
+
"learning_rate": 4.184565723615771e-05,
|
| 793 |
+
"loss": 0.1489630937576294,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.9379375318948484,
|
| 798 |
+
"grad_norm": 0.33672764897346497,
|
| 799 |
+
"learning_rate": 4.170638301967709e-05,
|
| 800 |
+
"loss": 0.2042783737182617,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.9463119741439096,
|
| 805 |
+
"grad_norm": 0.10248993337154388,
|
| 806 |
+
"learning_rate": 4.15661655684277e-05,
|
| 807 |
+
"loss": 0.1103521466255188,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.9546864163929707,
|
| 812 |
+
"grad_norm": 0.0915706530213356,
|
| 813 |
+
"learning_rate": 4.142501279902293e-05,
|
| 814 |
+
"loss": 0.16539056301116944,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.9630608586420318,
|
| 819 |
+
"grad_norm": 1.0003262758255005,
|
| 820 |
+
"learning_rate": 4.128293268088377e-05,
|
| 821 |
+
"loss": 0.2568183898925781,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.971435300891093,
|
| 826 |
+
"grad_norm": 0.21115422248840332,
|
| 827 |
+
"learning_rate": 4.1139933235788933e-05,
|
| 828 |
+
"loss": 0.14126954078674317,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.9798097431401541,
|
| 833 |
+
"grad_norm": 0.13956555724143982,
|
| 834 |
+
"learning_rate": 4.099602253742189e-05,
|
| 835 |
+
"loss": 0.17241095304489135,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.9881841853892153,
|
| 840 |
+
"grad_norm": 0.6704546809196472,
|
| 841 |
+
"learning_rate": 4.085120871091506e-05,
|
| 842 |
+
"loss": 0.12498955726623535,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.9965586276382764,
|
| 847 |
+
"grad_norm": 1.1212393045425415,
|
| 848 |
+
"learning_rate": 4.070549993239106e-05,
|
| 849 |
+
"loss": 0.16118084192276,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.0046059432369836,
|
| 854 |
+
"grad_norm": 2.63454008102417,
|
| 855 |
+
"learning_rate": 4.0558904428501107e-05,
|
| 856 |
+
"loss": 0.15824345350265503,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.0129803854860449,
|
| 861 |
+
"grad_norm": 0.6368555426597595,
|
| 862 |
+
"learning_rate": 4.041143047596051e-05,
|
| 863 |
+
"loss": 0.09336324334144593,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.0213548277351059,
|
| 868 |
+
"grad_norm": 0.3028055429458618,
|
| 869 |
+
"learning_rate": 4.026308640108139e-05,
|
| 870 |
+
"loss": 0.07328135371208191,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.0297292699841671,
|
| 875 |
+
"grad_norm": 1.0523555278778076,
|
| 876 |
+
"learning_rate": 4.01138805793026e-05,
|
| 877 |
+
"loss": 0.10240118503570557,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.0381037122332282,
|
| 882 |
+
"grad_norm": 0.669131875038147,
|
| 883 |
+
"learning_rate": 3.99638214347168e-05,
|
| 884 |
+
"loss": 0.060682100057601926,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.0464781544822894,
|
| 889 |
+
"grad_norm": 0.30694952607154846,
|
| 890 |
+
"learning_rate": 3.981291743959488e-05,
|
| 891 |
+
"loss": 0.13537111282348632,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.0548525967313505,
|
| 896 |
+
"grad_norm": 0.2553592026233673,
|
| 897 |
+
"learning_rate": 3.9661177113907625e-05,
|
| 898 |
+
"loss": 0.058023715019226076,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.0632270389804117,
|
| 903 |
+
"grad_norm": 1.2412865161895752,
|
| 904 |
+
"learning_rate": 3.950860902484461e-05,
|
| 905 |
+
"loss": 0.10125534534454346,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.0716014812294727,
|
| 910 |
+
"grad_norm": 0.34957438707351685,
|
| 911 |
+
"learning_rate": 3.935522178633062e-05,
|
| 912 |
+
"loss": 0.12363550662994385,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.079975923478534,
|
| 917 |
+
"grad_norm": 0.0916418582201004,
|
| 918 |
+
"learning_rate": 3.92010240585392e-05,
|
| 919 |
+
"loss": 0.11654759645462036,
|
| 920 |
+
"step": 2580
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.088350365727595,
|
| 924 |
+
"grad_norm": 1.0168888568878174,
|
| 925 |
+
"learning_rate": 3.904602454740376e-05,
|
| 926 |
+
"loss": 0.09794877171516418,
|
| 927 |
+
"step": 2600
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.0967248079766563,
|
| 931 |
+
"grad_norm": 1.4686931371688843,
|
| 932 |
+
"learning_rate": 3.889023200412603e-05,
|
| 933 |
+
"loss": 0.09772626161575318,
|
| 934 |
+
"step": 2620
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.1050992502257173,
|
| 938 |
+
"grad_norm": 0.42878222465515137,
|
| 939 |
+
"learning_rate": 3.873365522468198e-05,
|
| 940 |
+
"loss": 0.08697801232337951,
|
| 941 |
+
"step": 2640
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.1134736924747786,
|
| 945 |
+
"grad_norm": 0.6559142470359802,
|
| 946 |
+
"learning_rate": 3.85763030493252e-05,
|
| 947 |
+
"loss": 0.10625931024551391,
|
| 948 |
+
"step": 2660
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.1218481347238396,
|
| 952 |
+
"grad_norm": 0.4922764301300049,
|
| 953 |
+
"learning_rate": 3.841818436208776e-05,
|
| 954 |
+
"loss": 0.06333786249160767,
|
| 955 |
+
"step": 2680
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.1302225769729008,
|
| 959 |
+
"grad_norm": 0.09760922938585281,
|
| 960 |
+
"learning_rate": 3.825930809027866e-05,
|
| 961 |
+
"loss": 0.08181596398353577,
|
| 962 |
+
"step": 2700
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.1385970192219619,
|
| 966 |
+
"grad_norm": 0.10643378645181656,
|
| 967 |
+
"learning_rate": 3.809968320397979e-05,
|
| 968 |
+
"loss": 0.09863247275352478,
|
| 969 |
+
"step": 2720
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.1469714614710231,
|
| 973 |
+
"grad_norm": 0.12689609825611115,
|
| 974 |
+
"learning_rate": 3.793931871553942e-05,
|
| 975 |
+
"loss": 0.060498178005218506,
|
| 976 |
+
"step": 2740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.1553459037200842,
|
| 980 |
+
"grad_norm": 0.23766513168811798,
|
| 981 |
+
"learning_rate": 3.777822367906349e-05,
|
| 982 |
+
"loss": 0.06649444103240967,
|
| 983 |
+
"step": 2760
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.1637203459691454,
|
| 987 |
+
"grad_norm": 0.1511031687259674,
|
| 988 |
+
"learning_rate": 3.761640718990429e-05,
|
| 989 |
+
"loss": 0.07478836178779602,
|
| 990 |
+
"step": 2780
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.1720947882182067,
|
| 994 |
+
"grad_norm": 2.1615374088287354,
|
| 995 |
+
"learning_rate": 3.745387838414706e-05,
|
| 996 |
+
"loss": 0.18305596113204955,
|
| 997 |
+
"step": 2800
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.1804692304672677,
|
| 1001 |
+
"grad_norm": 0.11615941673517227,
|
| 1002 |
+
"learning_rate": 3.729064643809404e-05,
|
| 1003 |
+
"loss": 0.08897812366485595,
|
| 1004 |
+
"step": 2820
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.1888436727163287,
|
| 1008 |
+
"grad_norm": 0.21777822077274323,
|
| 1009 |
+
"learning_rate": 3.712672056774649e-05,
|
| 1010 |
+
"loss": 0.06442444324493408,
|
| 1011 |
+
"step": 2840
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.19721811496539,
|
| 1015 |
+
"grad_norm": 0.7141409516334534,
|
| 1016 |
+
"learning_rate": 3.696211002828434e-05,
|
| 1017 |
+
"loss": 0.09018256664276122,
|
| 1018 |
+
"step": 2860
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.2055925572144512,
|
| 1022 |
+
"grad_norm": 0.42543089389801025,
|
| 1023 |
+
"learning_rate": 3.679682411354358e-05,
|
| 1024 |
+
"loss": 0.09642559885978699,
|
| 1025 |
+
"step": 2880
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.2139669994635123,
|
| 1029 |
+
"grad_norm": 0.07673882693052292,
|
| 1030 |
+
"learning_rate": 3.6630872155491605e-05,
|
| 1031 |
+
"loss": 0.05595778226852417,
|
| 1032 |
+
"step": 2900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.2223414417125735,
|
| 1036 |
+
"grad_norm": 1.069671630859375,
|
| 1037 |
+
"learning_rate": 3.6464263523700306e-05,
|
| 1038 |
+
"loss": 0.0820774793624878,
|
| 1039 |
+
"step": 2920
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.2307158839616346,
|
| 1043 |
+
"grad_norm": 0.39548397064208984,
|
| 1044 |
+
"learning_rate": 3.6297007624817087e-05,
|
| 1045 |
+
"loss": 0.07778416872024536,
|
| 1046 |
+
"step": 2940
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.2390903262106958,
|
| 1050 |
+
"grad_norm": 0.39791959524154663,
|
| 1051 |
+
"learning_rate": 3.612911390203374e-05,
|
| 1052 |
+
"loss": 0.0765472412109375,
|
| 1053 |
+
"step": 2960
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.2474647684597568,
|
| 1057 |
+
"grad_norm": 0.5976597666740417,
|
| 1058 |
+
"learning_rate": 3.59605918345533e-05,
|
| 1059 |
+
"loss": 0.07529214024543762,
|
| 1060 |
+
"step": 2980
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.255839210708818,
|
| 1064 |
+
"grad_norm": 0.11279038339853287,
|
| 1065 |
+
"learning_rate": 3.5791450937054885e-05,
|
| 1066 |
+
"loss": 0.1001011610031128,
|
| 1067 |
+
"step": 3000
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.2642136529578791,
|
| 1071 |
+
"grad_norm": 1.1798434257507324,
|
| 1072 |
+
"learning_rate": 3.5621700759156435e-05,
|
| 1073 |
+
"loss": 0.09131324887275696,
|
| 1074 |
+
"step": 3020
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.2725880952069404,
|
| 1078 |
+
"grad_norm": 0.1061301976442337,
|
| 1079 |
+
"learning_rate": 3.545135088487562e-05,
|
| 1080 |
+
"loss": 0.08855420351028442,
|
| 1081 |
+
"step": 3040
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.2809625374560014,
|
| 1085 |
+
"grad_norm": 0.49484983086586,
|
| 1086 |
+
"learning_rate": 3.528041093208868e-05,
|
| 1087 |
+
"loss": 0.1239747166633606,
|
| 1088 |
+
"step": 3060
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.2893369797050627,
|
| 1092 |
+
"grad_norm": 1.3041439056396484,
|
| 1093 |
+
"learning_rate": 3.510889055198741e-05,
|
| 1094 |
+
"loss": 0.09958173632621765,
|
| 1095 |
+
"step": 3080
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.2977114219541237,
|
| 1099 |
+
"grad_norm": 1.169832468032837,
|
| 1100 |
+
"learning_rate": 3.493679942853427e-05,
|
| 1101 |
+
"loss": 0.0866806983947754,
|
| 1102 |
+
"step": 3100
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.306085864203185,
|
| 1106 |
+
"grad_norm": 0.6045485138893127,
|
| 1107 |
+
"learning_rate": 3.476414727791563e-05,
|
| 1108 |
+
"loss": 0.08419098258018494,
|
| 1109 |
+
"step": 3120
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.314460306452246,
|
| 1113 |
+
"grad_norm": 0.04711831361055374,
|
| 1114 |
+
"learning_rate": 3.459094384799322e-05,
|
| 1115 |
+
"loss": 0.09029738903045655,
|
| 1116 |
+
"step": 3140
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.3228347487013072,
|
| 1120 |
+
"grad_norm": 0.11531050503253937,
|
| 1121 |
+
"learning_rate": 3.44171989177537e-05,
|
| 1122 |
+
"loss": 0.13222168684005736,
|
| 1123 |
+
"step": 3160
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.3312091909503683,
|
| 1127 |
+
"grad_norm": 1.0880193710327148,
|
| 1128 |
+
"learning_rate": 3.4242922296756655e-05,
|
| 1129 |
+
"loss": 0.06968819499015808,
|
| 1130 |
+
"step": 3180
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.3395836331994295,
|
| 1134 |
+
"grad_norm": 0.2005595564842224,
|
| 1135 |
+
"learning_rate": 3.406812382458062e-05,
|
| 1136 |
+
"loss": 0.06197818517684937,
|
| 1137 |
+
"step": 3200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.3479580754484906,
|
| 1141 |
+
"grad_norm": 0.2917577922344208,
|
| 1142 |
+
"learning_rate": 3.3892813370267663e-05,
|
| 1143 |
+
"loss": 0.15936510562896727,
|
| 1144 |
+
"step": 3220
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.3563325176975518,
|
| 1148 |
+
"grad_norm": 0.07237739115953445,
|
| 1149 |
+
"learning_rate": 3.37170008317661e-05,
|
| 1150 |
+
"loss": 0.07917057275772095,
|
| 1151 |
+
"step": 3240
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.3647069599466128,
|
| 1155 |
+
"grad_norm": 1.0094565153121948,
|
| 1156 |
+
"learning_rate": 3.354069613537171e-05,
|
| 1157 |
+
"loss": 0.07092810273170472,
|
| 1158 |
+
"step": 3260
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.373081402195674,
|
| 1162 |
+
"grad_norm": 0.0927204117178917,
|
| 1163 |
+
"learning_rate": 3.3363909235167265e-05,
|
| 1164 |
+
"loss": 0.06920674443244934,
|
| 1165 |
+
"step": 3280
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.3814558444447353,
|
| 1169 |
+
"grad_norm": 0.22500236332416534,
|
| 1170 |
+
"learning_rate": 3.318665011246056e-05,
|
| 1171 |
+
"loss": 0.06283442378044128,
|
| 1172 |
+
"step": 3300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.3898302866937964,
|
| 1176 |
+
"grad_norm": 0.12858115136623383,
|
| 1177 |
+
"learning_rate": 3.3008928775220835e-05,
|
| 1178 |
+
"loss": 0.059798961877822875,
|
| 1179 |
+
"step": 3320
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.3982047289428574,
|
| 1183 |
+
"grad_norm": 0.8109415173530579,
|
| 1184 |
+
"learning_rate": 3.283075525751379e-05,
|
| 1185 |
+
"loss": 0.1134151816368103,
|
| 1186 |
+
"step": 3340
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.4065791711919187,
|
| 1190 |
+
"grad_norm": 0.3329622745513916,
|
| 1191 |
+
"learning_rate": 3.2652139618934976e-05,
|
| 1192 |
+
"loss": 0.05868912935256958,
|
| 1193 |
+
"step": 3360
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.41495361344098,
|
| 1197 |
+
"grad_norm": 0.07076062262058258,
|
| 1198 |
+
"learning_rate": 3.247309194404194e-05,
|
| 1199 |
+
"loss": 0.05698283910751343,
|
| 1200 |
+
"step": 3380
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.423328055690041,
|
| 1204 |
+
"grad_norm": 0.7275592684745789,
|
| 1205 |
+
"learning_rate": 3.229362234178476e-05,
|
| 1206 |
+
"loss": 0.057073181867599486,
|
| 1207 |
+
"step": 3400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.431702497939102,
|
| 1211 |
+
"grad_norm": 0.6140939593315125,
|
| 1212 |
+
"learning_rate": 3.2113740944935367e-05,
|
| 1213 |
+
"loss": 0.08581722974777221,
|
| 1214 |
+
"step": 3420
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.4400769401881632,
|
| 1218 |
+
"grad_norm": 0.7452869415283203,
|
| 1219 |
+
"learning_rate": 3.193345790951542e-05,
|
| 1220 |
+
"loss": 0.08401907682418823,
|
| 1221 |
+
"step": 3440
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.4484513824372245,
|
| 1225 |
+
"grad_norm": 0.10415206104516983,
|
| 1226 |
+
"learning_rate": 3.1752783414222895e-05,
|
| 1227 |
+
"loss": 0.0695533037185669,
|
| 1228 |
+
"step": 3460
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.4568258246862855,
|
| 1232 |
+
"grad_norm": 0.1527501344680786,
|
| 1233 |
+
"learning_rate": 3.1571727659857445e-05,
|
| 1234 |
+
"loss": 0.049785712361335756,
|
| 1235 |
+
"step": 3480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.4652002669353466,
|
| 1239 |
+
"grad_norm": 0.6354191899299622,
|
| 1240 |
+
"learning_rate": 3.139030086874439e-05,
|
| 1241 |
+
"loss": 0.05135859847068787,
|
| 1242 |
+
"step": 3500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.4735747091844078,
|
| 1246 |
+
"grad_norm": 0.13222479820251465,
|
| 1247 |
+
"learning_rate": 3.120851328415768e-05,
|
| 1248 |
+
"loss": 0.07700438499450683,
|
| 1249 |
+
"step": 3520
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.481949151433469,
|
| 1253 |
+
"grad_norm": 0.05289041996002197,
|
| 1254 |
+
"learning_rate": 3.1026375169741435e-05,
|
| 1255 |
+
"loss": 0.09370012879371643,
|
| 1256 |
+
"step": 3540
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.49032359368253,
|
| 1260 |
+
"grad_norm": 0.07858635485172272,
|
| 1261 |
+
"learning_rate": 3.084389680893056e-05,
|
| 1262 |
+
"loss": 0.05941493511199951,
|
| 1263 |
+
"step": 3560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.4986980359315913,
|
| 1267 |
+
"grad_norm": 0.07863116264343262,
|
| 1268 |
+
"learning_rate": 3.066108850437012e-05,
|
| 1269 |
+
"loss": 0.057811629772186277,
|
| 1270 |
+
"step": 3580
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.5070724781806524,
|
| 1274 |
+
"grad_norm": 0.3188619911670685,
|
| 1275 |
+
"learning_rate": 3.047796057733364e-05,
|
| 1276 |
+
"loss": 0.043448212742805484,
|
| 1277 |
+
"step": 3600
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.5154469204297136,
|
| 1281 |
+
"grad_norm": 0.8982509970664978,
|
| 1282 |
+
"learning_rate": 3.029452336714041e-05,
|
| 1283 |
+
"loss": 0.04573395550251007,
|
| 1284 |
+
"step": 3620
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.5238213626787747,
|
| 1288 |
+
"grad_norm": 0.5027258396148682,
|
| 1289 |
+
"learning_rate": 3.0110787230571656e-05,
|
| 1290 |
+
"loss": 0.07429202795028686,
|
| 1291 |
+
"step": 3640
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.532195804927836,
|
| 1295 |
+
"grad_norm": 0.22799259424209595,
|
| 1296 |
+
"learning_rate": 2.9926762541285914e-05,
|
| 1297 |
+
"loss": 0.081844562292099,
|
| 1298 |
+
"step": 3660
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.540570247176897,
|
| 1302 |
+
"grad_norm": 0.106113962829113,
|
| 1303 |
+
"learning_rate": 2.974245968923322e-05,
|
| 1304 |
+
"loss": 0.0663108229637146,
|
| 1305 |
+
"step": 3680
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.5489446894259582,
|
| 1309 |
+
"grad_norm": 0.1487455815076828,
|
| 1310 |
+
"learning_rate": 2.955788908006859e-05,
|
| 1311 |
+
"loss": 0.0660822868347168,
|
| 1312 |
+
"step": 3700
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.5573191316750195,
|
| 1316 |
+
"grad_norm": 0.11356580257415771,
|
| 1317 |
+
"learning_rate": 2.9373061134564467e-05,
|
| 1318 |
+
"loss": 0.04001219272613525,
|
| 1319 |
+
"step": 3720
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.5656935739240805,
|
| 1323 |
+
"grad_norm": 0.32575735449790955,
|
| 1324 |
+
"learning_rate": 2.918798628802239e-05,
|
| 1325 |
+
"loss": 0.05586833357810974,
|
| 1326 |
+
"step": 3740
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.5740680161731415,
|
| 1330 |
+
"grad_norm": 0.2437354028224945,
|
| 1331 |
+
"learning_rate": 2.9002674989683814e-05,
|
| 1332 |
+
"loss": 0.030624213814735412,
|
| 1333 |
+
"step": 3760
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.5824424584222028,
|
| 1337 |
+
"grad_norm": 0.30267608165740967,
|
| 1338 |
+
"learning_rate": 2.8817137702140162e-05,
|
| 1339 |
+
"loss": 0.05758292078971863,
|
| 1340 |
+
"step": 3780
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.590816900671264,
|
| 1344 |
+
"grad_norm": 0.4417203366756439,
|
| 1345 |
+
"learning_rate": 2.8631384900742098e-05,
|
| 1346 |
+
"loss": 0.05091829895973206,
|
| 1347 |
+
"step": 3800
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.599191342920325,
|
| 1351 |
+
"grad_norm": 1.4709017276763916,
|
| 1352 |
+
"learning_rate": 2.844542707300812e-05,
|
| 1353 |
+
"loss": 0.05626246929168701,
|
| 1354 |
+
"step": 3820
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.607565785169386,
|
| 1358 |
+
"grad_norm": 0.13606201112270355,
|
| 1359 |
+
"learning_rate": 2.8259274718032375e-05,
|
| 1360 |
+
"loss": 0.05818360447883606,
|
| 1361 |
+
"step": 3840
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.6159402274184473,
|
| 1365 |
+
"grad_norm": 0.28740158677101135,
|
| 1366 |
+
"learning_rate": 2.8072938345891976e-05,
|
| 1367 |
+
"loss": 0.04518155753612518,
|
| 1368 |
+
"step": 3860
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.6243146696675086,
|
| 1372 |
+
"grad_norm": 0.1473538875579834,
|
| 1373 |
+
"learning_rate": 2.7886428477053538e-05,
|
| 1374 |
+
"loss": 0.08722549676895142,
|
| 1375 |
+
"step": 3880
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.6326891119165696,
|
| 1379 |
+
"grad_norm": 0.4489794969558716,
|
| 1380 |
+
"learning_rate": 2.7699755641779224e-05,
|
| 1381 |
+
"loss": 0.08287249207496643,
|
| 1382 |
+
"step": 3900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.6410635541656307,
|
| 1386 |
+
"grad_norm": 0.5043429732322693,
|
| 1387 |
+
"learning_rate": 2.751293037953221e-05,
|
| 1388 |
+
"loss": 0.038405871391296385,
|
| 1389 |
+
"step": 3920
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.649437996414692,
|
| 1393 |
+
"grad_norm": 0.5577481985092163,
|
| 1394 |
+
"learning_rate": 2.7325963238381625e-05,
|
| 1395 |
+
"loss": 0.040457093715667726,
|
| 1396 |
+
"step": 3940
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.6578124386637532,
|
| 1400 |
+
"grad_norm": 0.1733057051897049,
|
| 1401 |
+
"learning_rate": 2.7138864774407025e-05,
|
| 1402 |
+
"loss": 0.059363311529159545,
|
| 1403 |
+
"step": 3960
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.6661868809128142,
|
| 1407 |
+
"grad_norm": 0.46242690086364746,
|
| 1408 |
+
"learning_rate": 2.695164555110237e-05,
|
| 1409 |
+
"loss": 0.039838004112243655,
|
| 1410 |
+
"step": 3980
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.6745613231618752,
|
| 1414 |
+
"grad_norm": 0.319814532995224,
|
| 1415 |
+
"learning_rate": 2.676431613877968e-05,
|
| 1416 |
+
"loss": 0.034502464532852176,
|
| 1417 |
+
"step": 4000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.6829357654109365,
|
| 1421 |
+
"grad_norm": 0.09238389879465103,
|
| 1422 |
+
"learning_rate": 2.657688711397215e-05,
|
| 1423 |
+
"loss": 0.04835406839847565,
|
| 1424 |
+
"step": 4020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.6913102076599977,
|
| 1428 |
+
"grad_norm": 0.06494881957769394,
|
| 1429 |
+
"learning_rate": 2.6389369058837077e-05,
|
| 1430 |
+
"loss": 0.058844029903411865,
|
| 1431 |
+
"step": 4040
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.6996846499090588,
|
| 1435 |
+
"grad_norm": 0.052071474492549896,
|
| 1436 |
+
"learning_rate": 2.6201772560558358e-05,
|
| 1437 |
+
"loss": 0.0473739355802536,
|
| 1438 |
+
"step": 4060
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.7080590921581198,
|
| 1442 |
+
"grad_norm": 0.35652586817741394,
|
| 1443 |
+
"learning_rate": 2.6014108210748767e-05,
|
| 1444 |
+
"loss": 0.07051045894622802,
|
| 1445 |
+
"step": 4080
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.716433534407181,
|
| 1449 |
+
"grad_norm": 0.7592489719390869,
|
| 1450 |
+
"learning_rate": 2.5826386604851942e-05,
|
| 1451 |
+
"loss": 0.05508745312690735,
|
| 1452 |
+
"step": 4100
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.7248079766562423,
|
| 1456 |
+
"grad_norm": 0.037354808300733566,
|
| 1457 |
+
"learning_rate": 2.563861834154415e-05,
|
| 1458 |
+
"loss": 0.03927508890628815,
|
| 1459 |
+
"step": 4120
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.7331824189053033,
|
| 1463 |
+
"grad_norm": 0.3896387219429016,
|
| 1464 |
+
"learning_rate": 2.5450814022135954e-05,
|
| 1465 |
+
"loss": 0.07120077013969421,
|
| 1466 |
+
"step": 4140
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.7415568611543644,
|
| 1470 |
+
"grad_norm": 0.07390138506889343,
|
| 1471 |
+
"learning_rate": 2.5262984249973593e-05,
|
| 1472 |
+
"loss": 0.06470881700515747,
|
| 1473 |
+
"step": 4160
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.7499313034034256,
|
| 1477 |
+
"grad_norm": 0.6417694091796875,
|
| 1478 |
+
"learning_rate": 2.5075139629840377e-05,
|
| 1479 |
+
"loss": 0.10211913585662842,
|
| 1480 |
+
"step": 4180
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.7583057456524869,
|
| 1484 |
+
"grad_norm": 0.2673373818397522,
|
| 1485 |
+
"learning_rate": 2.488729076735791e-05,
|
| 1486 |
+
"loss": 0.05772486329078674,
|
| 1487 |
+
"step": 4200
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.7666801879015481,
|
| 1491 |
+
"grad_norm": 0.09736882895231247,
|
| 1492 |
+
"learning_rate": 2.469944826838734e-05,
|
| 1493 |
+
"loss": 0.06817198395729065,
|
| 1494 |
+
"step": 4220
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.7750546301506092,
|
| 1498 |
+
"grad_norm": 0.3496638536453247,
|
| 1499 |
+
"learning_rate": 2.4511622738430514e-05,
|
| 1500 |
+
"loss": 0.07389255762100219,
|
| 1501 |
+
"step": 4240
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.7834290723996702,
|
| 1505 |
+
"grad_norm": 0.1945846825838089,
|
| 1506 |
+
"learning_rate": 2.432382478203122e-05,
|
| 1507 |
+
"loss": 0.06225880980491638,
|
| 1508 |
+
"step": 4260
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.7918035146487314,
|
| 1512 |
+
"grad_norm": 0.2939809560775757,
|
| 1513 |
+
"learning_rate": 2.4136065002176453e-05,
|
| 1514 |
+
"loss": 0.041797152161598204,
|
| 1515 |
+
"step": 4280
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.8001779568977927,
|
| 1519 |
+
"grad_norm": 0.5871041417121887,
|
| 1520 |
+
"learning_rate": 2.39483539996978e-05,
|
| 1521 |
+
"loss": 0.06685131192207336,
|
| 1522 |
+
"step": 4300
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.8085523991468537,
|
| 1526 |
+
"grad_norm": 0.7149155735969543,
|
| 1527 |
+
"learning_rate": 2.376070237267287e-05,
|
| 1528 |
+
"loss": 0.06200284361839294,
|
| 1529 |
+
"step": 4320
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.8169268413959148,
|
| 1533 |
+
"grad_norm": 0.16763842105865479,
|
| 1534 |
+
"learning_rate": 2.3573120715826972e-05,
|
| 1535 |
+
"loss": 0.04558481574058533,
|
| 1536 |
+
"step": 4340
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.825301283644976,
|
| 1540 |
+
"grad_norm": 0.05273763835430145,
|
| 1541 |
+
"learning_rate": 2.338561961993494e-05,
|
| 1542 |
+
"loss": 0.03969838619232178,
|
| 1543 |
+
"step": 4360
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.8336757258940373,
|
| 1547 |
+
"grad_norm": 2.0518643856048584,
|
| 1548 |
+
"learning_rate": 2.3198209671223167e-05,
|
| 1549 |
+
"loss": 0.08448796272277832,
|
| 1550 |
+
"step": 4380
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.8420501681430983,
|
| 1554 |
+
"grad_norm": 0.2578012943267822,
|
| 1555 |
+
"learning_rate": 2.301090145077191e-05,
|
| 1556 |
+
"loss": 0.057272589206695555,
|
| 1557 |
+
"step": 4400
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.8504246103921593,
|
| 1561 |
+
"grad_norm": 1.0227431058883667,
|
| 1562 |
+
"learning_rate": 2.2823705533917894e-05,
|
| 1563 |
+
"loss": 0.05425040721893311,
|
| 1564 |
+
"step": 4420
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.8587990526412206,
|
| 1568 |
+
"grad_norm": 0.9780853390693665,
|
| 1569 |
+
"learning_rate": 2.263663248965725e-05,
|
| 1570 |
+
"loss": 0.04819593727588654,
|
| 1571 |
+
"step": 4440
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.8671734948902818,
|
| 1575 |
+
"grad_norm": 0.09778815507888794,
|
| 1576 |
+
"learning_rate": 2.2449692880048775e-05,
|
| 1577 |
+
"loss": 0.04756914675235748,
|
| 1578 |
+
"step": 4460
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.8755479371393429,
|
| 1582 |
+
"grad_norm": 0.5774021148681641,
|
| 1583 |
+
"learning_rate": 2.226289725961758e-05,
|
| 1584 |
+
"loss": 0.05137239694595337,
|
| 1585 |
+
"step": 4480
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.883922379388404,
|
| 1589 |
+
"grad_norm": 0.2773257791996002,
|
| 1590 |
+
"learning_rate": 2.2076256174759255e-05,
|
| 1591 |
+
"loss": 0.10701142549514771,
|
| 1592 |
+
"step": 4500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.8922968216374652,
|
| 1596 |
+
"grad_norm": 0.8629113435745239,
|
| 1597 |
+
"learning_rate": 2.1889780163144357e-05,
|
| 1598 |
+
"loss": 0.05790225863456726,
|
| 1599 |
+
"step": 4520
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.9006712638865264,
|
| 1603 |
+
"grad_norm": 0.4331710934638977,
|
| 1604 |
+
"learning_rate": 2.1703479753123464e-05,
|
| 1605 |
+
"loss": 0.08746187090873718,
|
| 1606 |
+
"step": 4540
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.9090457061355874,
|
| 1610 |
+
"grad_norm": 0.3421229124069214,
|
| 1611 |
+
"learning_rate": 2.1517365463132824e-05,
|
| 1612 |
+
"loss": 0.04949420392513275,
|
| 1613 |
+
"step": 4560
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.9174201483846485,
|
| 1617 |
+
"grad_norm": 0.8141835331916809,
|
| 1618 |
+
"learning_rate": 2.133144780110038e-05,
|
| 1619 |
+
"loss": 0.059812629222869874,
|
| 1620 |
+
"step": 4580
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.9257945906337097,
|
| 1624 |
+
"grad_norm": 0.5663272738456726,
|
| 1625 |
+
"learning_rate": 2.1145737263852573e-05,
|
| 1626 |
+
"loss": 0.0783354640007019,
|
| 1627 |
+
"step": 4600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.934169032882771,
|
| 1631 |
+
"grad_norm": 0.4589388966560364,
|
| 1632 |
+
"learning_rate": 2.0960244336521667e-05,
|
| 1633 |
+
"loss": 0.05254002809524536,
|
| 1634 |
+
"step": 4620
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.942543475131832,
|
| 1638 |
+
"grad_norm": 0.11917776614427567,
|
| 1639 |
+
"learning_rate": 2.0774979491953777e-05,
|
| 1640 |
+
"loss": 0.053898555040359494,
|
| 1641 |
+
"step": 4640
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.950917917380893,
|
| 1645 |
+
"grad_norm": 1.0947282314300537,
|
| 1646 |
+
"learning_rate": 2.058995319011756e-05,
|
| 1647 |
+
"loss": 0.04134993851184845,
|
| 1648 |
+
"step": 4660
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.9592923596299543,
|
| 1652 |
+
"grad_norm": 0.0579100176692009,
|
| 1653 |
+
"learning_rate": 2.0405175877513656e-05,
|
| 1654 |
+
"loss": 0.079537433385849,
|
| 1655 |
+
"step": 4680
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.9676668018790155,
|
| 1659 |
+
"grad_norm": 0.3768811821937561,
|
| 1660 |
+
"learning_rate": 2.0220657986584912e-05,
|
| 1661 |
+
"loss": 0.05231003165245056,
|
| 1662 |
+
"step": 4700
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.9760412441280766,
|
| 1666 |
+
"grad_norm": 0.04796705022454262,
|
| 1667 |
+
"learning_rate": 2.0036409935127316e-05,
|
| 1668 |
+
"loss": 0.055571871995925906,
|
| 1669 |
+
"step": 4720
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.9844156863771376,
|
| 1673 |
+
"grad_norm": 0.18154506385326385,
|
| 1674 |
+
"learning_rate": 1.9852442125701856e-05,
|
| 1675 |
+
"loss": 0.06901991367340088,
|
| 1676 |
+
"step": 4740
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.9927901286261989,
|
| 1680 |
+
"grad_norm": 0.05525406450033188,
|
| 1681 |
+
"learning_rate": 1.9668764945047167e-05,
|
| 1682 |
+
"loss": 0.029008516669273378,
|
| 1683 |
+
"step": 4760
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.000837444224906,
|
| 1687 |
+
"grad_norm": 0.3389388620853424,
|
| 1688 |
+
"learning_rate": 1.9485388763493153e-05,
|
| 1689 |
+
"loss": 0.03650375008583069,
|
| 1690 |
+
"step": 4780
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.009211886473967,
|
| 1694 |
+
"grad_norm": 0.03471681848168373,
|
| 1695 |
+
"learning_rate": 1.9302323934375406e-05,
|
| 1696 |
+
"loss": 0.03689234256744385,
|
| 1697 |
+
"step": 4800
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.0175863287230285,
|
| 1701 |
+
"grad_norm": 0.29095345735549927,
|
| 1702 |
+
"learning_rate": 1.911958079345072e-05,
|
| 1703 |
+
"loss": 0.02743149399757385,
|
| 1704 |
+
"step": 4820
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 2.0259607709720897,
|
| 1708 |
+
"grad_norm": 0.4494439661502838,
|
| 1709 |
+
"learning_rate": 1.8937169658313526e-05,
|
| 1710 |
+
"loss": 0.05301372408866882,
|
| 1711 |
+
"step": 4840
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.0343352132211505,
|
| 1715 |
+
"grad_norm": 0.0394783653318882,
|
| 1716 |
+
"learning_rate": 1.875510082781335e-05,
|
| 1717 |
+
"loss": 0.029765251278877258,
|
| 1718 |
+
"step": 4860
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.0427096554702118,
|
| 1722 |
+
"grad_norm": 0.5054375529289246,
|
| 1723 |
+
"learning_rate": 1.8573384581473346e-05,
|
| 1724 |
+
"loss": 0.03222983479499817,
|
| 1725 |
+
"step": 4880
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.051084097719273,
|
| 1729 |
+
"grad_norm": 0.06276942044496536,
|
| 1730 |
+
"learning_rate": 1.839203117890993e-05,
|
| 1731 |
+
"loss": 0.025779253244400023,
|
| 1732 |
+
"step": 4900
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.0594585399683343,
|
| 1736 |
+
"grad_norm": 0.35279664397239685,
|
| 1737 |
+
"learning_rate": 1.8211050859253558e-05,
|
| 1738 |
+
"loss": 0.03500397801399231,
|
| 1739 |
+
"step": 4920
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.067832982217395,
|
| 1743 |
+
"grad_norm": 0.4527212083339691,
|
| 1744 |
+
"learning_rate": 1.803045384057055e-05,
|
| 1745 |
+
"loss": 0.038686040043830874,
|
| 1746 |
+
"step": 4940
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 2.0762074244664563,
|
| 1750 |
+
"grad_norm": 0.045888010412454605,
|
| 1751 |
+
"learning_rate": 1.7850250319286245e-05,
|
| 1752 |
+
"loss": 0.04277970790863037,
|
| 1753 |
+
"step": 4960
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.0845818667155176,
|
| 1757 |
+
"grad_norm": 0.17555783689022064,
|
| 1758 |
+
"learning_rate": 1.7670450469609306e-05,
|
| 1759 |
+
"loss": 0.043071445822715757,
|
| 1760 |
+
"step": 4980
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.092956308964579,
|
| 1764 |
+
"grad_norm": 0.29859426617622375,
|
| 1765 |
+
"learning_rate": 1.7491064442957273e-05,
|
| 1766 |
+
"loss": 0.022787241637706755,
|
| 1767 |
+
"step": 5000
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.1013307512136397,
|
| 1771 |
+
"grad_norm": 0.11047807335853577,
|
| 1772 |
+
"learning_rate": 1.7312102367383426e-05,
|
| 1773 |
+
"loss": 0.031673601269721983,
|
| 1774 |
+
"step": 5020
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.109705193462701,
|
| 1778 |
+
"grad_norm": 0.26167428493499756,
|
| 1779 |
+
"learning_rate": 1.7133574347004953e-05,
|
| 1780 |
+
"loss": 0.05480382442474365,
|
| 1781 |
+
"step": 5040
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.118079635711762,
|
| 1785 |
+
"grad_norm": 0.047067124396562576,
|
| 1786 |
+
"learning_rate": 1.695549046143251e-05,
|
| 1787 |
+
"loss": 0.045348826050758365,
|
| 1788 |
+
"step": 5060
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.1264540779608234,
|
| 1792 |
+
"grad_norm": 0.47584113478660583,
|
| 1793 |
+
"learning_rate": 1.6777860765201082e-05,
|
| 1794 |
+
"loss": 0.03209327757358551,
|
| 1795 |
+
"step": 5080
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.1348285202098847,
|
| 1799 |
+
"grad_norm": 1.9363760948181152,
|
| 1800 |
+
"learning_rate": 1.6600695287202314e-05,
|
| 1801 |
+
"loss": 0.03423079252243042,
|
| 1802 |
+
"step": 5100
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.1432029624589455,
|
| 1806 |
+
"grad_norm": 0.7151720523834229,
|
| 1807 |
+
"learning_rate": 1.6424004030118355e-05,
|
| 1808 |
+
"loss": 0.03726245760917664,
|
| 1809 |
+
"step": 5120
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.1515774047080067,
|
| 1813 |
+
"grad_norm": 1.0615071058273315,
|
| 1814 |
+
"learning_rate": 1.624779696985701e-05,
|
| 1815 |
+
"loss": 0.02985242009162903,
|
| 1816 |
+
"step": 5140
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.159951846957068,
|
| 1820 |
+
"grad_norm": 0.08171194046735764,
|
| 1821 |
+
"learning_rate": 1.6072084054988563e-05,
|
| 1822 |
+
"loss": 0.07047683000564575,
|
| 1823 |
+
"step": 5160
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.1683262892061292,
|
| 1827 |
+
"grad_norm": 0.04042057693004608,
|
| 1828 |
+
"learning_rate": 1.5896875206184088e-05,
|
| 1829 |
+
"loss": 0.03392764925956726,
|
| 1830 |
+
"step": 5180
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.17670073145519,
|
| 1834 |
+
"grad_norm": 0.24893136322498322,
|
| 1835 |
+
"learning_rate": 1.5722180315655314e-05,
|
| 1836 |
+
"loss": 0.0692134976387024,
|
| 1837 |
+
"step": 5200
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.1850751737042513,
|
| 1841 |
+
"grad_norm": 0.5698428750038147,
|
| 1842 |
+
"learning_rate": 1.5548009246596108e-05,
|
| 1843 |
+
"loss": 0.03124482333660126,
|
| 1844 |
+
"step": 5220
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.1934496159533126,
|
| 1848 |
+
"grad_norm": 0.4285091757774353,
|
| 1849 |
+
"learning_rate": 1.5374371832625615e-05,
|
| 1850 |
+
"loss": 0.028138580918312072,
|
| 1851 |
+
"step": 5240
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.201824058202374,
|
| 1855 |
+
"grad_norm": 0.17444312572479248,
|
| 1856 |
+
"learning_rate": 1.5201277877233084e-05,
|
| 1857 |
+
"loss": 0.03015301823616028,
|
| 1858 |
+
"step": 5260
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.2101985004514346,
|
| 1862 |
+
"grad_norm": 0.05330653861165047,
|
| 1863 |
+
"learning_rate": 1.5028737153224303e-05,
|
| 1864 |
+
"loss": 0.02524380385875702,
|
| 1865 |
+
"step": 5280
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.218572942700496,
|
| 1869 |
+
"grad_norm": 0.35742461681365967,
|
| 1870 |
+
"learning_rate": 1.4856759402169901e-05,
|
| 1871 |
+
"loss": 0.03431325256824493,
|
| 1872 |
+
"step": 5300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.226947384949557,
|
| 1876 |
+
"grad_norm": 0.06930524855852127,
|
| 1877 |
+
"learning_rate": 1.4685354333855298e-05,
|
| 1878 |
+
"loss": 0.03083210587501526,
|
| 1879 |
+
"step": 5320
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.2353218271986184,
|
| 1883 |
+
"grad_norm": 0.07338620722293854,
|
| 1884 |
+
"learning_rate": 1.4514531625732498e-05,
|
| 1885 |
+
"loss": 0.04376775920391083,
|
| 1886 |
+
"step": 5340
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.243696269447679,
|
| 1890 |
+
"grad_norm": 0.4916664659976959,
|
| 1891 |
+
"learning_rate": 1.4344300922373727e-05,
|
| 1892 |
+
"loss": 0.038047611713409424,
|
| 1893 |
+
"step": 5360
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.2520707116967404,
|
| 1897 |
+
"grad_norm": 0.3985106945037842,
|
| 1898 |
+
"learning_rate": 1.4174671834926889e-05,
|
| 1899 |
+
"loss": 0.06458678245544433,
|
| 1900 |
+
"step": 5380
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.2604451539458017,
|
| 1904 |
+
"grad_norm": 0.38805675506591797,
|
| 1905 |
+
"learning_rate": 1.4005653940572922e-05,
|
| 1906 |
+
"loss": 0.024541012942790985,
|
| 1907 |
+
"step": 5400
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.268819596194863,
|
| 1911 |
+
"grad_norm": 0.03862249106168747,
|
| 1912 |
+
"learning_rate": 1.383725678198509e-05,
|
| 1913 |
+
"loss": 0.04991849362850189,
|
| 1914 |
+
"step": 5420
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.2771940384439238,
|
| 1918 |
+
"grad_norm": 0.12288820743560791,
|
| 1919 |
+
"learning_rate": 1.366948986679018e-05,
|
| 1920 |
+
"loss": 0.04350382089614868,
|
| 1921 |
+
"step": 5440
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.285568480692985,
|
| 1925 |
+
"grad_norm": 0.346986323595047,
|
| 1926 |
+
"learning_rate": 1.3502362667031759e-05,
|
| 1927 |
+
"loss": 0.037441790103912354,
|
| 1928 |
+
"step": 5460
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.2939429229420463,
|
| 1932 |
+
"grad_norm": 0.03941166400909424,
|
| 1933 |
+
"learning_rate": 1.333588461863532e-05,
|
| 1934 |
+
"loss": 0.02921281158924103,
|
| 1935 |
+
"step": 5480
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.3023173651911075,
|
| 1939 |
+
"grad_norm": 0.04610476642847061,
|
| 1940 |
+
"learning_rate": 1.317006512087558e-05,
|
| 1941 |
+
"loss": 0.02918812334537506,
|
| 1942 |
+
"step": 5500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.3106918074401683,
|
| 1946 |
+
"grad_norm": 0.08153572678565979,
|
| 1947 |
+
"learning_rate": 1.300491353584577e-05,
|
| 1948 |
+
"loss": 0.028163161873817445,
|
| 1949 |
+
"step": 5520
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.3190662496892296,
|
| 1953 |
+
"grad_norm": 0.03823208808898926,
|
| 1954 |
+
"learning_rate": 1.284043918792911e-05,
|
| 1955 |
+
"loss": 0.026886764168739318,
|
| 1956 |
+
"step": 5540
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.327440691938291,
|
| 1960 |
+
"grad_norm": 0.036283787339925766,
|
| 1961 |
+
"learning_rate": 1.267665136327229e-05,
|
| 1962 |
+
"loss": 0.03849519193172455,
|
| 1963 |
+
"step": 5560
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.335815134187352,
|
| 1967 |
+
"grad_norm": 0.36579447984695435,
|
| 1968 |
+
"learning_rate": 1.2513559309261208e-05,
|
| 1969 |
+
"loss": 0.029686197638511658,
|
| 1970 |
+
"step": 5580
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.3441895764364133,
|
| 1974 |
+
"grad_norm": 0.06716123968362808,
|
| 1975 |
+
"learning_rate": 1.2351172233998896e-05,
|
| 1976 |
+
"loss": 0.02747002840042114,
|
| 1977 |
+
"step": 5600
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.352564018685474,
|
| 1981 |
+
"grad_norm": 0.15424710512161255,
|
| 1982 |
+
"learning_rate": 1.2189499305785582e-05,
|
| 1983 |
+
"loss": 0.028761863708496094,
|
| 1984 |
+
"step": 5620
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.3609384609345354,
|
| 1988 |
+
"grad_norm": 0.23852872848510742,
|
| 1989 |
+
"learning_rate": 1.2028549652601077e-05,
|
| 1990 |
+
"loss": 0.024596422910690308,
|
| 1991 |
+
"step": 5640
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.3693129031835967,
|
| 1995 |
+
"grad_norm": 0.5796898603439331,
|
| 1996 |
+
"learning_rate": 1.1868332361589438e-05,
|
| 1997 |
+
"loss": 0.04195750653743744,
|
| 1998 |
+
"step": 5660
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.3776873454326575,
|
| 2002 |
+
"grad_norm": 0.5441113114356995,
|
| 2003 |
+
"learning_rate": 1.1708856478545865e-05,
|
| 2004 |
+
"loss": 0.026001009345054626,
|
| 2005 |
+
"step": 5680
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.3860617876817187,
|
| 2009 |
+
"grad_norm": 0.12801289558410645,
|
| 2010 |
+
"learning_rate": 1.1550131007406007e-05,
|
| 2011 |
+
"loss": 0.032627201080322264,
|
| 2012 |
+
"step": 5700
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.39443622993078,
|
| 2016 |
+
"grad_norm": 0.07165171205997467,
|
| 2017 |
+
"learning_rate": 1.1392164909737593e-05,
|
| 2018 |
+
"loss": 0.0330066978931427,
|
| 2019 |
+
"step": 5720
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.4028106721798412,
|
| 2023 |
+
"grad_norm": 0.06438246369361877,
|
| 2024 |
+
"learning_rate": 1.1234967104234495e-05,
|
| 2025 |
+
"loss": 0.041339311003685,
|
| 2026 |
+
"step": 5740
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.4111851144289025,
|
| 2030 |
+
"grad_norm": 0.3250308632850647,
|
| 2031 |
+
"learning_rate": 1.1078546466213141e-05,
|
| 2032 |
+
"loss": 0.0224463552236557,
|
| 2033 |
+
"step": 5760
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.4195595566779633,
|
| 2037 |
+
"grad_norm": 0.25930482149124146,
|
| 2038 |
+
"learning_rate": 1.0922911827111434e-05,
|
| 2039 |
+
"loss": 0.042743980884552,
|
| 2040 |
+
"step": 5780
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.4279339989270246,
|
| 2044 |
+
"grad_norm": 0.9659811854362488,
|
| 2045 |
+
"learning_rate": 1.0768071973990145e-05,
|
| 2046 |
+
"loss": 0.023346297442913055,
|
| 2047 |
+
"step": 5800
|
| 2048 |
+
}
|
| 2049 |
+
],
|
| 2050 |
+
"logging_steps": 20,
|
| 2051 |
+
"max_steps": 8362,
|
| 2052 |
+
"num_input_tokens_seen": 0,
|
| 2053 |
+
"num_train_epochs": 4,
|
| 2054 |
+
"save_steps": 1000000000,
|
| 2055 |
+
"stateful_callbacks": {
|
| 2056 |
+
"TrainerControl": {
|
| 2057 |
+
"args": {
|
| 2058 |
+
"should_epoch_stop": false,
|
| 2059 |
+
"should_evaluate": false,
|
| 2060 |
+
"should_log": false,
|
| 2061 |
+
"should_save": true,
|
| 2062 |
+
"should_training_stop": false
|
| 2063 |
+
},
|
| 2064 |
+
"attributes": {}
|
| 2065 |
+
}
|
| 2066 |
+
},
|
| 2067 |
+
"total_flos": 1.835656645169664e+16,
|
| 2068 |
+
"train_batch_size": 1,
|
| 2069 |
+
"trial_name": null,
|
| 2070 |
+
"trial_params": null
|
| 2071 |
+
}
|