Training in progress, step 4000, checkpoint
Browse files- last-checkpoint/config.json +33 -0
- last-checkpoint/generation_config.json +9 -0
- last-checkpoint/model.safetensors +3 -0
- last-checkpoint/optimizer.pt +3 -0
- last-checkpoint/rng_state.pth +3 -0
- last-checkpoint/scheduler.pt +3 -0
- last-checkpoint/trainer_state.json +3154 -0
- last-checkpoint/training_args.bin +3 -0
last-checkpoint/config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"HybridTimeScaleLM"
|
| 4 |
+
],
|
| 5 |
+
"bos_token_id": 1,
|
| 6 |
+
"dropout": 0.05,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"latent_dim": 768,
|
| 10 |
+
"layer_types": [
|
| 11 |
+
"linear",
|
| 12 |
+
"linear",
|
| 13 |
+
"linear",
|
| 14 |
+
"softmax",
|
| 15 |
+
"linear",
|
| 16 |
+
"linear",
|
| 17 |
+
"linear",
|
| 18 |
+
"softmax",
|
| 19 |
+
"linear",
|
| 20 |
+
"linear",
|
| 21 |
+
"linear",
|
| 22 |
+
"softmax"
|
| 23 |
+
],
|
| 24 |
+
"model_type": "hybrid_fourier_lm",
|
| 25 |
+
"num_layers": 12,
|
| 26 |
+
"num_modes": 64,
|
| 27 |
+
"pad_token_id": 2,
|
| 28 |
+
"tie_word_embeddings": true,
|
| 29 |
+
"time_scale": 128.0,
|
| 30 |
+
"transformers_version": "5.13.1",
|
| 31 |
+
"use_cache": false,
|
| 32 |
+
"vocab_size": 32768
|
| 33 |
+
}
|
last-checkpoint/generation_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"output_attentions": false,
|
| 6 |
+
"output_hidden_states": false,
|
| 7 |
+
"pad_token_id": 2,
|
| 8 |
+
"transformers_version": "5.13.1"
|
| 9 |
+
}
|
last-checkpoint/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef40685adde7bab73c68f2cde8502c190c2320aaeb5986766f734c5533fdeea4
|
| 3 |
+
size 570000656
|
last-checkpoint/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eda448ad879f83a12acd08ff28ab243a118f3be634cc8549ca759dc7db0a3f41
|
| 3 |
+
size 1622091
|
last-checkpoint/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8bb9979b869aa629dc0b87f8cba3b2ce182a3a887a6304833dc30df5b8fd0e2
|
| 3 |
+
size 14455
|
last-checkpoint/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccc95881d63aeb232a0d233120305836db2087b7d53cc84df6ba9d3559169ce0
|
| 3 |
+
size 1465
|
last-checkpoint/trainer_state.json
ADDED
|
@@ -0,0 +1,3154 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.36363636363636365,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 4000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0009090909090909091,
|
| 14 |
+
"grad_norm": Infinity,
|
| 15 |
+
"learning_rate": 9.999983482695944e-07,
|
| 16 |
+
"loss": 2.972582244873047,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0018181818181818182,
|
| 21 |
+
"grad_norm": Infinity,
|
| 22 |
+
"learning_rate": 9.999926385982519e-07,
|
| 23 |
+
"loss": 3.030682182312012,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0027272727272727275,
|
| 28 |
+
"grad_norm": Infinity,
|
| 29 |
+
"learning_rate": 9.999828506408003e-07,
|
| 30 |
+
"loss": 2.846811294555664,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0036363636363636364,
|
| 35 |
+
"grad_norm": 1.1991735607408722e+19,
|
| 36 |
+
"learning_rate": 9.999689844770766e-07,
|
| 37 |
+
"loss": 2.7316921234130858,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.004545454545454545,
|
| 42 |
+
"grad_norm": 6.335976986745242e+18,
|
| 43 |
+
"learning_rate": 9.999510402201835e-07,
|
| 44 |
+
"loss": 2.9967422485351562,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.005454545454545455,
|
| 49 |
+
"grad_norm": Infinity,
|
| 50 |
+
"learning_rate": 9.99929018016486e-07,
|
| 51 |
+
"loss": 2.680373191833496,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.006363636363636364,
|
| 56 |
+
"grad_norm": Infinity,
|
| 57 |
+
"learning_rate": 9.999029180456129e-07,
|
| 58 |
+
"loss": 2.49072265625,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.007272727272727273,
|
| 63 |
+
"grad_norm": 1.1419123145802514e+19,
|
| 64 |
+
"learning_rate": 9.998727405204534e-07,
|
| 65 |
+
"loss": 2.3953088760375976,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.008181818181818182,
|
| 70 |
+
"grad_norm": Infinity,
|
| 71 |
+
"learning_rate": 9.998384856871564e-07,
|
| 72 |
+
"loss": 2.656139945983887,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.00909090909090909,
|
| 77 |
+
"grad_norm": Infinity,
|
| 78 |
+
"learning_rate": 9.998001538251282e-07,
|
| 79 |
+
"loss": 2.8163694381713866,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.00909090909090909,
|
| 84 |
+
"eval_loss": 2.7871742248535156,
|
| 85 |
+
"eval_runtime": 25.2604,
|
| 86 |
+
"eval_samples_per_second": 5.819,
|
| 87 |
+
"eval_steps_per_second": 1.465,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.01,
|
| 92 |
+
"grad_norm": Infinity,
|
| 93 |
+
"learning_rate": 9.997577452470298e-07,
|
| 94 |
+
"loss": 2.816670608520508,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.01090909090909091,
|
| 99 |
+
"grad_norm": 1.794573070979249e+19,
|
| 100 |
+
"learning_rate": 9.99711260298775e-07,
|
| 101 |
+
"loss": 2.6952537536621093,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.011818181818181818,
|
| 106 |
+
"grad_norm": Infinity,
|
| 107 |
+
"learning_rate": 9.99660699359527e-07,
|
| 108 |
+
"loss": 2.715743064880371,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.012727272727272728,
|
| 113 |
+
"grad_norm": Infinity,
|
| 114 |
+
"learning_rate": 9.996060628416963e-07,
|
| 115 |
+
"loss": 3.068762016296387,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.013636363636363636,
|
| 120 |
+
"grad_norm": Infinity,
|
| 121 |
+
"learning_rate": 9.995473511909357e-07,
|
| 122 |
+
"loss": 2.851385498046875,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.014545454545454545,
|
| 127 |
+
"grad_norm": Infinity,
|
| 128 |
+
"learning_rate": 9.994845648861382e-07,
|
| 129 |
+
"loss": 2.8277969360351562,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.015454545454545455,
|
| 134 |
+
"grad_norm": Infinity,
|
| 135 |
+
"learning_rate": 9.994177044394322e-07,
|
| 136 |
+
"loss": 2.8090538024902343,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.016363636363636365,
|
| 141 |
+
"grad_norm": 5.987306456414683e+18,
|
| 142 |
+
"learning_rate": 9.993467703961784e-07,
|
| 143 |
+
"loss": 2.7782918930053713,
|
| 144 |
+
"step": 180
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.017272727272727273,
|
| 148 |
+
"grad_norm": Infinity,
|
| 149 |
+
"learning_rate": 9.992717633349635e-07,
|
| 150 |
+
"loss": 2.7075550079345705,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.01818181818181818,
|
| 155 |
+
"grad_norm": Infinity,
|
| 156 |
+
"learning_rate": 9.99192683867597e-07,
|
| 157 |
+
"loss": 3.050520133972168,
|
| 158 |
+
"step": 200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.01818181818181818,
|
| 162 |
+
"eval_loss": 2.787334680557251,
|
| 163 |
+
"eval_runtime": 25.3811,
|
| 164 |
+
"eval_samples_per_second": 5.792,
|
| 165 |
+
"eval_steps_per_second": 1.458,
|
| 166 |
+
"step": 200
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.019090909090909092,
|
| 170 |
+
"grad_norm": 2.708298899595985e+18,
|
| 171 |
+
"learning_rate": 9.99109532639106e-07,
|
| 172 |
+
"loss": 2.5547176361083985,
|
| 173 |
+
"step": 210
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.02,
|
| 177 |
+
"grad_norm": Infinity,
|
| 178 |
+
"learning_rate": 9.99022310327729e-07,
|
| 179 |
+
"loss": 2.71762638092041,
|
| 180 |
+
"step": 220
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.02090909090909091,
|
| 184 |
+
"grad_norm": 1.4361367883219468e+19,
|
| 185 |
+
"learning_rate": 9.98931017644912e-07,
|
| 186 |
+
"loss": 2.555472564697266,
|
| 187 |
+
"step": 230
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.02181818181818182,
|
| 191 |
+
"grad_norm": Infinity,
|
| 192 |
+
"learning_rate": 9.988356553353017e-07,
|
| 193 |
+
"loss": 3.093941497802734,
|
| 194 |
+
"step": 240
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.022727272727272728,
|
| 198 |
+
"grad_norm": Infinity,
|
| 199 |
+
"learning_rate": 9.987362241767383e-07,
|
| 200 |
+
"loss": 3.011051368713379,
|
| 201 |
+
"step": 250
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.023636363636363636,
|
| 205 |
+
"grad_norm": 5.583390414590706e+18,
|
| 206 |
+
"learning_rate": 9.986327249802515e-07,
|
| 207 |
+
"loss": 2.736302947998047,
|
| 208 |
+
"step": 260
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.024545454545454544,
|
| 212 |
+
"grad_norm": 1.3582739927916872e+19,
|
| 213 |
+
"learning_rate": 9.985251585900525e-07,
|
| 214 |
+
"loss": 2.630130577087402,
|
| 215 |
+
"step": 270
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.025454545454545455,
|
| 219 |
+
"grad_norm": 4.937351466969989e+18,
|
| 220 |
+
"learning_rate": 9.984135258835272e-07,
|
| 221 |
+
"loss": 2.6608613967895507,
|
| 222 |
+
"step": 280
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.026363636363636363,
|
| 226 |
+
"grad_norm": 1.5744615083612832e+19,
|
| 227 |
+
"learning_rate": 9.98297827771229e-07,
|
| 228 |
+
"loss": 2.79409236907959,
|
| 229 |
+
"step": 290
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.02727272727272727,
|
| 233 |
+
"grad_norm": Infinity,
|
| 234 |
+
"learning_rate": 9.981780651968722e-07,
|
| 235 |
+
"loss": 2.7820377349853516,
|
| 236 |
+
"step": 300
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.02727272727272727,
|
| 240 |
+
"eval_loss": 2.7874202728271484,
|
| 241 |
+
"eval_runtime": 25.2656,
|
| 242 |
+
"eval_samples_per_second": 5.818,
|
| 243 |
+
"eval_steps_per_second": 1.464,
|
| 244 |
+
"step": 300
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.028181818181818183,
|
| 248 |
+
"grad_norm": 5.150588553037611e+18,
|
| 249 |
+
"learning_rate": 9.980542391373232e-07,
|
| 250 |
+
"loss": 2.862852668762207,
|
| 251 |
+
"step": 310
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.02909090909090909,
|
| 255 |
+
"grad_norm": Infinity,
|
| 256 |
+
"learning_rate": 9.979263506025929e-07,
|
| 257 |
+
"loss": 2.890887451171875,
|
| 258 |
+
"step": 320
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.03,
|
| 262 |
+
"grad_norm": Infinity,
|
| 263 |
+
"learning_rate": 9.977944006358285e-07,
|
| 264 |
+
"loss": 2.572690963745117,
|
| 265 |
+
"step": 330
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.03090909090909091,
|
| 269 |
+
"grad_norm": Infinity,
|
| 270 |
+
"learning_rate": 9.976583903133059e-07,
|
| 271 |
+
"loss": 2.8408248901367186,
|
| 272 |
+
"step": 340
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.031818181818181815,
|
| 276 |
+
"grad_norm": Infinity,
|
| 277 |
+
"learning_rate": 9.975183207444189e-07,
|
| 278 |
+
"loss": 2.8312393188476563,
|
| 279 |
+
"step": 350
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.03272727272727273,
|
| 283 |
+
"grad_norm": 5.449467699304333e+18,
|
| 284 |
+
"learning_rate": 9.97374193071672e-07,
|
| 285 |
+
"loss": 3.0894535064697264,
|
| 286 |
+
"step": 360
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.03363636363636364,
|
| 290 |
+
"grad_norm": Infinity,
|
| 291 |
+
"learning_rate": 9.972260084706704e-07,
|
| 292 |
+
"loss": 2.8833297729492187,
|
| 293 |
+
"step": 370
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.034545454545454546,
|
| 297 |
+
"grad_norm": 5.602140936134984e+18,
|
| 298 |
+
"learning_rate": 9.970737681501104e-07,
|
| 299 |
+
"loss": 3.0639154434204103,
|
| 300 |
+
"step": 380
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.035454545454545454,
|
| 304 |
+
"grad_norm": 6.436255195977482e+18,
|
| 305 |
+
"learning_rate": 9.969174733517693e-07,
|
| 306 |
+
"loss": 2.7005983352661134,
|
| 307 |
+
"step": 390
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.03636363636363636,
|
| 311 |
+
"grad_norm": 6.883443067668398e+18,
|
| 312 |
+
"learning_rate": 9.967571253504956e-07,
|
| 313 |
+
"loss": 2.7406898498535157,
|
| 314 |
+
"step": 400
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.03636363636363636,
|
| 318 |
+
"eval_loss": 2.787506580352783,
|
| 319 |
+
"eval_runtime": 25.3731,
|
| 320 |
+
"eval_samples_per_second": 5.794,
|
| 321 |
+
"eval_steps_per_second": 1.458,
|
| 322 |
+
"step": 400
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.03727272727272727,
|
| 326 |
+
"grad_norm": Infinity,
|
| 327 |
+
"learning_rate": 9.965927254541986e-07,
|
| 328 |
+
"loss": 3.2558399200439454,
|
| 329 |
+
"step": 410
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.038181818181818185,
|
| 333 |
+
"grad_norm": Infinity,
|
| 334 |
+
"learning_rate": 9.964242750038379e-07,
|
| 335 |
+
"loss": 3.5612926483154297,
|
| 336 |
+
"step": 420
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.03909090909090909,
|
| 340 |
+
"grad_norm": Infinity,
|
| 341 |
+
"learning_rate": 9.96251775373412e-07,
|
| 342 |
+
"loss": 2.661521339416504,
|
| 343 |
+
"step": 430
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 0.04,
|
| 347 |
+
"grad_norm": Infinity,
|
| 348 |
+
"learning_rate": 9.960752279699469e-07,
|
| 349 |
+
"loss": 2.7075674057006838,
|
| 350 |
+
"step": 440
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.04090909090909091,
|
| 354 |
+
"grad_norm": 1.3481015311138292e+19,
|
| 355 |
+
"learning_rate": 9.958946342334858e-07,
|
| 356 |
+
"loss": 2.752682685852051,
|
| 357 |
+
"step": 450
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.04181818181818182,
|
| 361 |
+
"grad_norm": Infinity,
|
| 362 |
+
"learning_rate": 9.957099956370761e-07,
|
| 363 |
+
"loss": 3.051430892944336,
|
| 364 |
+
"step": 460
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.042727272727272725,
|
| 368 |
+
"grad_norm": Infinity,
|
| 369 |
+
"learning_rate": 9.95521313686758e-07,
|
| 370 |
+
"loss": 2.702523040771484,
|
| 371 |
+
"step": 470
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.04363636363636364,
|
| 375 |
+
"grad_norm": Infinity,
|
| 376 |
+
"learning_rate": 9.953285899215524e-07,
|
| 377 |
+
"loss": 2.804890251159668,
|
| 378 |
+
"step": 480
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.04454545454545455,
|
| 382 |
+
"grad_norm": Infinity,
|
| 383 |
+
"learning_rate": 9.951318259134473e-07,
|
| 384 |
+
"loss": 2.9871782302856444,
|
| 385 |
+
"step": 490
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.045454545454545456,
|
| 389 |
+
"grad_norm": Infinity,
|
| 390 |
+
"learning_rate": 9.949310232673867e-07,
|
| 391 |
+
"loss": 2.5105640411376955,
|
| 392 |
+
"step": 500
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.045454545454545456,
|
| 396 |
+
"eval_loss": 2.7876086235046387,
|
| 397 |
+
"eval_runtime": 25.1494,
|
| 398 |
+
"eval_samples_per_second": 5.845,
|
| 399 |
+
"eval_steps_per_second": 1.471,
|
| 400 |
+
"step": 500
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.046363636363636364,
|
| 404 |
+
"grad_norm": Infinity,
|
| 405 |
+
"learning_rate": 9.947261836212555e-07,
|
| 406 |
+
"loss": 2.7869237899780273,
|
| 407 |
+
"step": 510
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.04727272727272727,
|
| 411 |
+
"grad_norm": Infinity,
|
| 412 |
+
"learning_rate": 9.94517308645868e-07,
|
| 413 |
+
"loss": 2.9804574966430666,
|
| 414 |
+
"step": 520
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.04818181818181818,
|
| 418 |
+
"grad_norm": Infinity,
|
| 419 |
+
"learning_rate": 9.94304400044953e-07,
|
| 420 |
+
"loss": 2.777148628234863,
|
| 421 |
+
"step": 530
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 0.04909090909090909,
|
| 425 |
+
"grad_norm": Infinity,
|
| 426 |
+
"learning_rate": 9.940874595551404e-07,
|
| 427 |
+
"loss": 2.6402297973632813,
|
| 428 |
+
"step": 540
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.05,
|
| 432 |
+
"grad_norm": 6.372751802403652e+18,
|
| 433 |
+
"learning_rate": 9.938664889459468e-07,
|
| 434 |
+
"loss": 2.805499267578125,
|
| 435 |
+
"step": 550
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.05090909090909091,
|
| 439 |
+
"grad_norm": Infinity,
|
| 440 |
+
"learning_rate": 9.936414900197618e-07,
|
| 441 |
+
"loss": 2.7490556716918944,
|
| 442 |
+
"step": 560
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.05181818181818182,
|
| 446 |
+
"grad_norm": Infinity,
|
| 447 |
+
"learning_rate": 9.934124646118324e-07,
|
| 448 |
+
"loss": 2.7342456817626952,
|
| 449 |
+
"step": 570
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.05272727272727273,
|
| 453 |
+
"grad_norm": Infinity,
|
| 454 |
+
"learning_rate": 9.931794145902485e-07,
|
| 455 |
+
"loss": 3.1899335861206053,
|
| 456 |
+
"step": 580
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.053636363636363635,
|
| 460 |
+
"grad_norm": Infinity,
|
| 461 |
+
"learning_rate": 9.92942341855927e-07,
|
| 462 |
+
"loss": 3.0976799011230467,
|
| 463 |
+
"step": 590
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.05454545454545454,
|
| 467 |
+
"grad_norm": 5.517231700435796e+18,
|
| 468 |
+
"learning_rate": 9.927012483425976e-07,
|
| 469 |
+
"loss": 2.8919891357421874,
|
| 470 |
+
"step": 600
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.05454545454545454,
|
| 474 |
+
"eval_loss": 2.7876956462860107,
|
| 475 |
+
"eval_runtime": 25.3974,
|
| 476 |
+
"eval_samples_per_second": 5.788,
|
| 477 |
+
"eval_steps_per_second": 1.457,
|
| 478 |
+
"step": 600
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.05545454545454546,
|
| 482 |
+
"grad_norm": Infinity,
|
| 483 |
+
"learning_rate": 9.924561360167858e-07,
|
| 484 |
+
"loss": 2.845229721069336,
|
| 485 |
+
"step": 610
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.056363636363636366,
|
| 489 |
+
"grad_norm": Infinity,
|
| 490 |
+
"learning_rate": 9.922070068777972e-07,
|
| 491 |
+
"loss": 2.4370473861694335,
|
| 492 |
+
"step": 620
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.057272727272727274,
|
| 496 |
+
"grad_norm": Infinity,
|
| 497 |
+
"learning_rate": 9.91953862957702e-07,
|
| 498 |
+
"loss": 2.5011491775512695,
|
| 499 |
+
"step": 630
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.05818181818181818,
|
| 503 |
+
"grad_norm": 1.8244524918785966e+18,
|
| 504 |
+
"learning_rate": 9.91696706321317e-07,
|
| 505 |
+
"loss": 3.2636795043945312,
|
| 506 |
+
"step": 640
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.05909090909090909,
|
| 510 |
+
"grad_norm": Infinity,
|
| 511 |
+
"learning_rate": 9.914355390661895e-07,
|
| 512 |
+
"loss": 2.698811340332031,
|
| 513 |
+
"step": 650
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.06,
|
| 517 |
+
"grad_norm": 5.180378721080574e+18,
|
| 518 |
+
"learning_rate": 9.91170363322581e-07,
|
| 519 |
+
"loss": 2.5908447265625,
|
| 520 |
+
"step": 660
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.060909090909090906,
|
| 524 |
+
"grad_norm": 2.5083164262012027e+18,
|
| 525 |
+
"learning_rate": 9.90901181253448e-07,
|
| 526 |
+
"loss": 2.9120254516601562,
|
| 527 |
+
"step": 670
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.06181818181818182,
|
| 531 |
+
"grad_norm": Infinity,
|
| 532 |
+
"learning_rate": 9.906279950544258e-07,
|
| 533 |
+
"loss": 2.828919219970703,
|
| 534 |
+
"step": 680
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.06272727272727273,
|
| 538 |
+
"grad_norm": Infinity,
|
| 539 |
+
"learning_rate": 9.903508069538107e-07,
|
| 540 |
+
"loss": 3.0443912506103517,
|
| 541 |
+
"step": 690
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.06363636363636363,
|
| 545 |
+
"grad_norm": Infinity,
|
| 546 |
+
"learning_rate": 9.900696192125401e-07,
|
| 547 |
+
"loss": 2.940337562561035,
|
| 548 |
+
"step": 700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.06363636363636363,
|
| 552 |
+
"eval_loss": 2.787841796875,
|
| 553 |
+
"eval_runtime": 25.3047,
|
| 554 |
+
"eval_samples_per_second": 5.809,
|
| 555 |
+
"eval_steps_per_second": 1.462,
|
| 556 |
+
"step": 700
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.06454545454545454,
|
| 560 |
+
"grad_norm": Infinity,
|
| 561 |
+
"learning_rate": 9.897844341241764e-07,
|
| 562 |
+
"loss": 2.8803565979003904,
|
| 563 |
+
"step": 710
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.06545454545454546,
|
| 567 |
+
"grad_norm": Infinity,
|
| 568 |
+
"learning_rate": 9.894952540148861e-07,
|
| 569 |
+
"loss": 2.391953468322754,
|
| 570 |
+
"step": 720
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.06636363636363636,
|
| 574 |
+
"grad_norm": 7.620495189789901e+18,
|
| 575 |
+
"learning_rate": 9.892020812434227e-07,
|
| 576 |
+
"loss": 2.8717506408691404,
|
| 577 |
+
"step": 730
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.06727272727272728,
|
| 581 |
+
"grad_norm": Infinity,
|
| 582 |
+
"learning_rate": 9.889049182011064e-07,
|
| 583 |
+
"loss": 2.985699462890625,
|
| 584 |
+
"step": 740
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.06818181818181818,
|
| 588 |
+
"grad_norm": Infinity,
|
| 589 |
+
"learning_rate": 9.886037673118046e-07,
|
| 590 |
+
"loss": 2.728943634033203,
|
| 591 |
+
"step": 750
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.06909090909090909,
|
| 595 |
+
"grad_norm": Infinity,
|
| 596 |
+
"learning_rate": 9.882986310319123e-07,
|
| 597 |
+
"loss": 2.5520145416259767,
|
| 598 |
+
"step": 760
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.07,
|
| 602 |
+
"grad_norm": 3.4245235227083407e+18,
|
| 603 |
+
"learning_rate": 9.879895118503322e-07,
|
| 604 |
+
"loss": 2.822847366333008,
|
| 605 |
+
"step": 770
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.07090909090909091,
|
| 609 |
+
"grad_norm": Infinity,
|
| 610 |
+
"learning_rate": 9.876764122884546e-07,
|
| 611 |
+
"loss": 2.851523780822754,
|
| 612 |
+
"step": 780
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.07181818181818182,
|
| 616 |
+
"grad_norm": Infinity,
|
| 617 |
+
"learning_rate": 9.873593349001363e-07,
|
| 618 |
+
"loss": 2.833907699584961,
|
| 619 |
+
"step": 790
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.07272727272727272,
|
| 623 |
+
"grad_norm": 5.103755405008306e+18,
|
| 624 |
+
"learning_rate": 9.870382822716797e-07,
|
| 625 |
+
"loss": 2.9371898651123045,
|
| 626 |
+
"step": 800
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.07272727272727272,
|
| 630 |
+
"eval_loss": 2.7879319190979004,
|
| 631 |
+
"eval_runtime": 25.258,
|
| 632 |
+
"eval_samples_per_second": 5.82,
|
| 633 |
+
"eval_steps_per_second": 1.465,
|
| 634 |
+
"step": 800
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.07363636363636364,
|
| 638 |
+
"grad_norm": Infinity,
|
| 639 |
+
"learning_rate": 9.867132570218127e-07,
|
| 640 |
+
"loss": 2.694431114196777,
|
| 641 |
+
"step": 810
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.07454545454545454,
|
| 645 |
+
"grad_norm": Infinity,
|
| 646 |
+
"learning_rate": 9.863842618016658e-07,
|
| 647 |
+
"loss": 3.0608633041381834,
|
| 648 |
+
"step": 820
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.07545454545454545,
|
| 652 |
+
"grad_norm": Infinity,
|
| 653 |
+
"learning_rate": 9.86051299294752e-07,
|
| 654 |
+
"loss": 2.6543277740478515,
|
| 655 |
+
"step": 830
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.07636363636363637,
|
| 659 |
+
"grad_norm": 1.7660894526035722e+19,
|
| 660 |
+
"learning_rate": 9.857143722169442e-07,
|
| 661 |
+
"loss": 2.9908905029296875,
|
| 662 |
+
"step": 840
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.07727272727272727,
|
| 666 |
+
"grad_norm": Infinity,
|
| 667 |
+
"learning_rate": 9.853734833164525e-07,
|
| 668 |
+
"loss": 2.996399688720703,
|
| 669 |
+
"step": 850
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.07818181818181819,
|
| 673 |
+
"grad_norm": Infinity,
|
| 674 |
+
"learning_rate": 9.850286353738032e-07,
|
| 675 |
+
"loss": 2.7943418502807615,
|
| 676 |
+
"step": 860
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.07909090909090909,
|
| 680 |
+
"grad_norm": Infinity,
|
| 681 |
+
"learning_rate": 9.846798312018146e-07,
|
| 682 |
+
"loss": 2.3392452239990233,
|
| 683 |
+
"step": 870
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.08,
|
| 687 |
+
"grad_norm": Infinity,
|
| 688 |
+
"learning_rate": 9.843270736455752e-07,
|
| 689 |
+
"loss": 2.653145980834961,
|
| 690 |
+
"step": 880
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.0809090909090909,
|
| 694 |
+
"grad_norm": Infinity,
|
| 695 |
+
"learning_rate": 9.839703655824194e-07,
|
| 696 |
+
"loss": 2.9171478271484377,
|
| 697 |
+
"step": 890
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.08181818181818182,
|
| 701 |
+
"grad_norm": Infinity,
|
| 702 |
+
"learning_rate": 9.836097099219057e-07,
|
| 703 |
+
"loss": 2.7383825302124025,
|
| 704 |
+
"step": 900
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.08181818181818182,
|
| 708 |
+
"eval_loss": 2.7882080078125,
|
| 709 |
+
"eval_runtime": 25.34,
|
| 710 |
+
"eval_samples_per_second": 5.801,
|
| 711 |
+
"eval_steps_per_second": 1.46,
|
| 712 |
+
"step": 900
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 0.08272727272727273,
|
| 716 |
+
"grad_norm": Infinity,
|
| 717 |
+
"learning_rate": 9.832451096057912e-07,
|
| 718 |
+
"loss": 2.408688545227051,
|
| 719 |
+
"step": 910
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.08363636363636363,
|
| 723 |
+
"grad_norm": 1.1698251764699496e+19,
|
| 724 |
+
"learning_rate": 9.82876567608008e-07,
|
| 725 |
+
"loss": 2.5717914581298826,
|
| 726 |
+
"step": 920
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.08454545454545455,
|
| 730 |
+
"grad_norm": Infinity,
|
| 731 |
+
"learning_rate": 9.825040869346404e-07,
|
| 732 |
+
"loss": 2.71414737701416,
|
| 733 |
+
"step": 930
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.08545454545454545,
|
| 737 |
+
"grad_norm": Infinity,
|
| 738 |
+
"learning_rate": 9.821276706238985e-07,
|
| 739 |
+
"loss": 2.639026069641113,
|
| 740 |
+
"step": 940
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.08636363636363636,
|
| 744 |
+
"grad_norm": Infinity,
|
| 745 |
+
"learning_rate": 9.81747321746094e-07,
|
| 746 |
+
"loss": 2.969831848144531,
|
| 747 |
+
"step": 950
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.08727272727272728,
|
| 751 |
+
"grad_norm": 4.923415706843742e+18,
|
| 752 |
+
"learning_rate": 9.81363043403616e-07,
|
| 753 |
+
"loss": 2.8209333419799805,
|
| 754 |
+
"step": 960
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.08818181818181818,
|
| 758 |
+
"grad_norm": Infinity,
|
| 759 |
+
"learning_rate": 9.809748387309047e-07,
|
| 760 |
+
"loss": 2.7035985946655274,
|
| 761 |
+
"step": 970
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.0890909090909091,
|
| 765 |
+
"grad_norm": Infinity,
|
| 766 |
+
"learning_rate": 9.805827108944259e-07,
|
| 767 |
+
"loss": 3.101240348815918,
|
| 768 |
+
"step": 980
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.09,
|
| 772 |
+
"grad_norm": Infinity,
|
| 773 |
+
"learning_rate": 9.801866630926459e-07,
|
| 774 |
+
"loss": 3.0513349533081056,
|
| 775 |
+
"step": 990
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.09090909090909091,
|
| 779 |
+
"grad_norm": Infinity,
|
| 780 |
+
"learning_rate": 9.797866985560049e-07,
|
| 781 |
+
"loss": 3.109476852416992,
|
| 782 |
+
"step": 1000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.09090909090909091,
|
| 786 |
+
"eval_loss": 2.7881357669830322,
|
| 787 |
+
"eval_runtime": 25.2713,
|
| 788 |
+
"eval_samples_per_second": 5.817,
|
| 789 |
+
"eval_steps_per_second": 1.464,
|
| 790 |
+
"step": 1000
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 0.09181818181818181,
|
| 794 |
+
"grad_norm": 8.761301725526098e+18,
|
| 795 |
+
"learning_rate": 9.793828205468902e-07,
|
| 796 |
+
"loss": 2.617064666748047,
|
| 797 |
+
"step": 1010
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 0.09272727272727273,
|
| 801 |
+
"grad_norm": Infinity,
|
| 802 |
+
"learning_rate": 9.789750323596107e-07,
|
| 803 |
+
"loss": 2.809076118469238,
|
| 804 |
+
"step": 1020
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.09363636363636364,
|
| 808 |
+
"grad_norm": Infinity,
|
| 809 |
+
"learning_rate": 9.78563337320369e-07,
|
| 810 |
+
"loss": 2.4070892333984375,
|
| 811 |
+
"step": 1030
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 0.09454545454545454,
|
| 815 |
+
"grad_norm": Infinity,
|
| 816 |
+
"learning_rate": 9.781477387872351e-07,
|
| 817 |
+
"loss": 2.7776512145996093,
|
| 818 |
+
"step": 1040
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.09545454545454546,
|
| 822 |
+
"grad_norm": 1.8563528976131686e+18,
|
| 823 |
+
"learning_rate": 9.77728240150118e-07,
|
| 824 |
+
"loss": 2.613321876525879,
|
| 825 |
+
"step": 1050
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.09636363636363636,
|
| 829 |
+
"grad_norm": Infinity,
|
| 830 |
+
"learning_rate": 9.773048448307396e-07,
|
| 831 |
+
"loss": 2.667472076416016,
|
| 832 |
+
"step": 1060
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 0.09727272727272727,
|
| 836 |
+
"grad_norm": Infinity,
|
| 837 |
+
"learning_rate": 9.768775562826047e-07,
|
| 838 |
+
"loss": 2.8902267456054687,
|
| 839 |
+
"step": 1070
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 0.09818181818181818,
|
| 843 |
+
"grad_norm": Infinity,
|
| 844 |
+
"learning_rate": 9.764463779909748e-07,
|
| 845 |
+
"loss": 3.081578826904297,
|
| 846 |
+
"step": 1080
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 0.09909090909090909,
|
| 850 |
+
"grad_norm": Infinity,
|
| 851 |
+
"learning_rate": 9.760113134728383e-07,
|
| 852 |
+
"loss": 2.976548194885254,
|
| 853 |
+
"step": 1090
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 0.1,
|
| 857 |
+
"grad_norm": Infinity,
|
| 858 |
+
"learning_rate": 9.755723662768827e-07,
|
| 859 |
+
"loss": 2.6694435119628905,
|
| 860 |
+
"step": 1100
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.1,
|
| 864 |
+
"eval_loss": 2.7883572578430176,
|
| 865 |
+
"eval_runtime": 25.3715,
|
| 866 |
+
"eval_samples_per_second": 5.794,
|
| 867 |
+
"eval_steps_per_second": 1.458,
|
| 868 |
+
"step": 1100
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 0.1009090909090909,
|
| 872 |
+
"grad_norm": Infinity,
|
| 873 |
+
"learning_rate": 9.751295399834653e-07,
|
| 874 |
+
"loss": 3.10753173828125,
|
| 875 |
+
"step": 1110
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 0.10181818181818182,
|
| 879 |
+
"grad_norm": 1.1276669319796228e+19,
|
| 880 |
+
"learning_rate": 9.746828382045843e-07,
|
| 881 |
+
"loss": 3.002843475341797,
|
| 882 |
+
"step": 1120
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.10272727272727272,
|
| 886 |
+
"grad_norm": Infinity,
|
| 887 |
+
"learning_rate": 9.742322645838476e-07,
|
| 888 |
+
"loss": 3.1718584060668946,
|
| 889 |
+
"step": 1130
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.10363636363636364,
|
| 893 |
+
"grad_norm": Infinity,
|
| 894 |
+
"learning_rate": 9.737778227964468e-07,
|
| 895 |
+
"loss": 2.706940460205078,
|
| 896 |
+
"step": 1140
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.10454545454545454,
|
| 900 |
+
"grad_norm": Infinity,
|
| 901 |
+
"learning_rate": 9.73319516549123e-07,
|
| 902 |
+
"loss": 3.022201156616211,
|
| 903 |
+
"step": 1150
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.10545454545454545,
|
| 907 |
+
"grad_norm": Infinity,
|
| 908 |
+
"learning_rate": 9.728573495801392e-07,
|
| 909 |
+
"loss": 2.8677652359008787,
|
| 910 |
+
"step": 1160
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.10636363636363637,
|
| 914 |
+
"grad_norm": Infinity,
|
| 915 |
+
"learning_rate": 9.723913256592495e-07,
|
| 916 |
+
"loss": 2.7335657119750976,
|
| 917 |
+
"step": 1170
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.10727272727272727,
|
| 921 |
+
"grad_norm": Infinity,
|
| 922 |
+
"learning_rate": 9.719214485876675e-07,
|
| 923 |
+
"loss": 2.5799468994140624,
|
| 924 |
+
"step": 1180
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.10818181818181818,
|
| 928 |
+
"grad_norm": Infinity,
|
| 929 |
+
"learning_rate": 9.71447722198036e-07,
|
| 930 |
+
"loss": 2.8072345733642576,
|
| 931 |
+
"step": 1190
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 0.10909090909090909,
|
| 935 |
+
"grad_norm": 4.2832869091002286e+18,
|
| 936 |
+
"learning_rate": 9.709701503543952e-07,
|
| 937 |
+
"loss": 2.8874731063842773,
|
| 938 |
+
"step": 1200
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.10909090909090909,
|
| 942 |
+
"eval_loss": 2.788444757461548,
|
| 943 |
+
"eval_runtime": 25.4151,
|
| 944 |
+
"eval_samples_per_second": 5.784,
|
| 945 |
+
"eval_steps_per_second": 1.456,
|
| 946 |
+
"step": 1200
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 0.11,
|
| 950 |
+
"grad_norm": Infinity,
|
| 951 |
+
"learning_rate": 9.70488736952152e-07,
|
| 952 |
+
"loss": 2.6035261154174805,
|
| 953 |
+
"step": 1210
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.11090909090909092,
|
| 957 |
+
"grad_norm": Infinity,
|
| 958 |
+
"learning_rate": 9.70003485918047e-07,
|
| 959 |
+
"loss": 3.0014669418334963,
|
| 960 |
+
"step": 1220
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.11181818181818182,
|
| 964 |
+
"grad_norm": 1.4744464122615693e+19,
|
| 965 |
+
"learning_rate": 9.695144012101237e-07,
|
| 966 |
+
"loss": 3.6563724517822265,
|
| 967 |
+
"step": 1230
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 0.11272727272727273,
|
| 971 |
+
"grad_norm": Infinity,
|
| 972 |
+
"learning_rate": 9.690214868176957e-07,
|
| 973 |
+
"loss": 3.150925636291504,
|
| 974 |
+
"step": 1240
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.11363636363636363,
|
| 978 |
+
"grad_norm": 1.3018001069077168e+19,
|
| 979 |
+
"learning_rate": 9.68524746761314e-07,
|
| 980 |
+
"loss": 2.9263046264648436,
|
| 981 |
+
"step": 1250
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.11454545454545455,
|
| 985 |
+
"grad_norm": Infinity,
|
| 986 |
+
"learning_rate": 9.680241850927343e-07,
|
| 987 |
+
"loss": 2.8352964401245115,
|
| 988 |
+
"step": 1260
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 0.11545454545454545,
|
| 992 |
+
"grad_norm": 3.9445782289952276e+18,
|
| 993 |
+
"learning_rate": 9.675198058948842e-07,
|
| 994 |
+
"loss": 2.9046945571899414,
|
| 995 |
+
"step": 1270
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.11636363636363636,
|
| 999 |
+
"grad_norm": Infinity,
|
| 1000 |
+
"learning_rate": 9.670116132818296e-07,
|
| 1001 |
+
"loss": 2.561818504333496,
|
| 1002 |
+
"step": 1280
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 0.11727272727272728,
|
| 1006 |
+
"grad_norm": Infinity,
|
| 1007 |
+
"learning_rate": 9.664996113987413e-07,
|
| 1008 |
+
"loss": 2.5337194442749023,
|
| 1009 |
+
"step": 1290
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 0.11818181818181818,
|
| 1013 |
+
"grad_norm": Infinity,
|
| 1014 |
+
"learning_rate": 9.659838044218612e-07,
|
| 1015 |
+
"loss": 2.7506519317626954,
|
| 1016 |
+
"step": 1300
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.11818181818181818,
|
| 1020 |
+
"eval_loss": 2.7886366844177246,
|
| 1021 |
+
"eval_runtime": 25.2984,
|
| 1022 |
+
"eval_samples_per_second": 5.811,
|
| 1023 |
+
"eval_steps_per_second": 1.463,
|
| 1024 |
+
"step": 1300
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.1190909090909091,
|
| 1028 |
+
"grad_norm": Infinity,
|
| 1029 |
+
"learning_rate": 9.654641965584677e-07,
|
| 1030 |
+
"loss": 2.585481643676758,
|
| 1031 |
+
"step": 1310
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.12,
|
| 1035 |
+
"grad_norm": Infinity,
|
| 1036 |
+
"learning_rate": 9.649407920468429e-07,
|
| 1037 |
+
"loss": 2.387790489196777,
|
| 1038 |
+
"step": 1320
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.12090909090909091,
|
| 1042 |
+
"grad_norm": 8.597805445987434e+18,
|
| 1043 |
+
"learning_rate": 9.644135951562358e-07,
|
| 1044 |
+
"loss": 2.409588623046875,
|
| 1045 |
+
"step": 1330
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.12181818181818181,
|
| 1049 |
+
"grad_norm": 1.4816577791746638e+19,
|
| 1050 |
+
"learning_rate": 9.638826101868297e-07,
|
| 1051 |
+
"loss": 2.848589324951172,
|
| 1052 |
+
"step": 1340
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.12272727272727273,
|
| 1056 |
+
"grad_norm": Infinity,
|
| 1057 |
+
"learning_rate": 9.63347841469705e-07,
|
| 1058 |
+
"loss": 2.937419891357422,
|
| 1059 |
+
"step": 1350
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.12363636363636364,
|
| 1063 |
+
"grad_norm": Infinity,
|
| 1064 |
+
"learning_rate": 9.62809293366806e-07,
|
| 1065 |
+
"loss": 2.8799938201904296,
|
| 1066 |
+
"step": 1360
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.12454545454545454,
|
| 1070 |
+
"grad_norm": Infinity,
|
| 1071 |
+
"learning_rate": 9.62266970270904e-07,
|
| 1072 |
+
"loss": 2.6749975204467775,
|
| 1073 |
+
"step": 1370
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.12545454545454546,
|
| 1077 |
+
"grad_norm": 1.299667714056808e+19,
|
| 1078 |
+
"learning_rate": 9.617208766055612e-07,
|
| 1079 |
+
"loss": 2.8736942291259764,
|
| 1080 |
+
"step": 1380
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.12636363636363637,
|
| 1084 |
+
"grad_norm": Infinity,
|
| 1085 |
+
"learning_rate": 9.61171016825096e-07,
|
| 1086 |
+
"loss": 2.868055534362793,
|
| 1087 |
+
"step": 1390
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.12727272727272726,
|
| 1091 |
+
"grad_norm": 6.926160743674937e+18,
|
| 1092 |
+
"learning_rate": 9.606173954145452e-07,
|
| 1093 |
+
"loss": 3.0848974227905273,
|
| 1094 |
+
"step": 1400
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.12727272727272726,
|
| 1098 |
+
"eval_loss": 2.7885942459106445,
|
| 1099 |
+
"eval_runtime": 25.4457,
|
| 1100 |
+
"eval_samples_per_second": 5.777,
|
| 1101 |
+
"eval_steps_per_second": 1.454,
|
| 1102 |
+
"step": 1400
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.12818181818181817,
|
| 1106 |
+
"grad_norm": Infinity,
|
| 1107 |
+
"learning_rate": 9.600600168896283e-07,
|
| 1108 |
+
"loss": 2.7365055084228516,
|
| 1109 |
+
"step": 1410
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.1290909090909091,
|
| 1113 |
+
"grad_norm": Infinity,
|
| 1114 |
+
"learning_rate": 9.594988857967105e-07,
|
| 1115 |
+
"loss": 3.1326847076416016,
|
| 1116 |
+
"step": 1420
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.13,
|
| 1120 |
+
"grad_norm": Infinity,
|
| 1121 |
+
"learning_rate": 9.589340067127657e-07,
|
| 1122 |
+
"loss": 2.2127132415771484,
|
| 1123 |
+
"step": 1430
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.13090909090909092,
|
| 1127 |
+
"grad_norm": Infinity,
|
| 1128 |
+
"learning_rate": 9.583653842453382e-07,
|
| 1129 |
+
"loss": 2.7018590927124024,
|
| 1130 |
+
"step": 1440
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.1318181818181818,
|
| 1134 |
+
"grad_norm": Infinity,
|
| 1135 |
+
"learning_rate": 9.577930230325072e-07,
|
| 1136 |
+
"loss": 3.1343578338623046,
|
| 1137 |
+
"step": 1450
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.13272727272727272,
|
| 1141 |
+
"grad_norm": Infinity,
|
| 1142 |
+
"learning_rate": 9.572169277428464e-07,
|
| 1143 |
+
"loss": 2.7940521240234375,
|
| 1144 |
+
"step": 1460
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.13363636363636364,
|
| 1148 |
+
"grad_norm": Infinity,
|
| 1149 |
+
"learning_rate": 9.566371030753882e-07,
|
| 1150 |
+
"loss": 2.700839614868164,
|
| 1151 |
+
"step": 1470
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.13454545454545455,
|
| 1155 |
+
"grad_norm": Infinity,
|
| 1156 |
+
"learning_rate": 9.560535537595838e-07,
|
| 1157 |
+
"loss": 2.729141616821289,
|
| 1158 |
+
"step": 1480
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.13545454545454547,
|
| 1162 |
+
"grad_norm": 9.33644581265526e+18,
|
| 1163 |
+
"learning_rate": 9.554662845552656e-07,
|
| 1164 |
+
"loss": 2.583717155456543,
|
| 1165 |
+
"step": 1490
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.13636363636363635,
|
| 1169 |
+
"grad_norm": Infinity,
|
| 1170 |
+
"learning_rate": 9.548753002526076e-07,
|
| 1171 |
+
"loss": 2.8652286529541016,
|
| 1172 |
+
"step": 1500
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.13636363636363635,
|
| 1176 |
+
"eval_loss": 2.7888333797454834,
|
| 1177 |
+
"eval_runtime": 25.5103,
|
| 1178 |
+
"eval_samples_per_second": 5.762,
|
| 1179 |
+
"eval_steps_per_second": 1.45,
|
| 1180 |
+
"step": 1500
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 0.13727272727272727,
|
| 1184 |
+
"grad_norm": Infinity,
|
| 1185 |
+
"learning_rate": 9.54280605672087e-07,
|
| 1186 |
+
"loss": 2.878238868713379,
|
| 1187 |
+
"step": 1510
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 0.13818181818181818,
|
| 1191 |
+
"grad_norm": Infinity,
|
| 1192 |
+
"learning_rate": 9.536822056644444e-07,
|
| 1193 |
+
"loss": 2.555144500732422,
|
| 1194 |
+
"step": 1520
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.1390909090909091,
|
| 1198 |
+
"grad_norm": Infinity,
|
| 1199 |
+
"learning_rate": 9.530801051106448e-07,
|
| 1200 |
+
"loss": 2.8380672454833986,
|
| 1201 |
+
"step": 1530
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 0.14,
|
| 1205 |
+
"grad_norm": 1.8191689261902725e+19,
|
| 1206 |
+
"learning_rate": 9.52474308921837e-07,
|
| 1207 |
+
"loss": 2.9135501861572264,
|
| 1208 |
+
"step": 1540
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.1409090909090909,
|
| 1212 |
+
"grad_norm": Infinity,
|
| 1213 |
+
"learning_rate": 9.518648220393142e-07,
|
| 1214 |
+
"loss": 2.995115280151367,
|
| 1215 |
+
"step": 1550
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.14181818181818182,
|
| 1219 |
+
"grad_norm": Infinity,
|
| 1220 |
+
"learning_rate": 9.512516494344732e-07,
|
| 1221 |
+
"loss": 2.3885705947875975,
|
| 1222 |
+
"step": 1560
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.14272727272727273,
|
| 1226 |
+
"grad_norm": Infinity,
|
| 1227 |
+
"learning_rate": 9.506347961087743e-07,
|
| 1228 |
+
"loss": 3.084286689758301,
|
| 1229 |
+
"step": 1570
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.14363636363636365,
|
| 1233 |
+
"grad_norm": Infinity,
|
| 1234 |
+
"learning_rate": 9.500142670937006e-07,
|
| 1235 |
+
"loss": 2.691249656677246,
|
| 1236 |
+
"step": 1580
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 0.14454545454545453,
|
| 1240 |
+
"grad_norm": Infinity,
|
| 1241 |
+
"learning_rate": 9.493900674507158e-07,
|
| 1242 |
+
"loss": 2.8905223846435546,
|
| 1243 |
+
"step": 1590
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 0.14545454545454545,
|
| 1247 |
+
"grad_norm": 1.9084444600023122e+18,
|
| 1248 |
+
"learning_rate": 9.487622022712247e-07,
|
| 1249 |
+
"loss": 2.9005517959594727,
|
| 1250 |
+
"step": 1600
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.14545454545454545,
|
| 1254 |
+
"eval_loss": 2.7890355587005615,
|
| 1255 |
+
"eval_runtime": 25.4165,
|
| 1256 |
+
"eval_samples_per_second": 5.784,
|
| 1257 |
+
"eval_steps_per_second": 1.456,
|
| 1258 |
+
"step": 1600
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.14636363636363636,
|
| 1262 |
+
"grad_norm": Infinity,
|
| 1263 |
+
"learning_rate": 9.481306766765304e-07,
|
| 1264 |
+
"loss": 2.458705520629883,
|
| 1265 |
+
"step": 1610
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.14727272727272728,
|
| 1269 |
+
"grad_norm": 7.697958532745789e+18,
|
| 1270 |
+
"learning_rate": 9.474954958177925e-07,
|
| 1271 |
+
"loss": 2.9042383193969727,
|
| 1272 |
+
"step": 1620
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.1481818181818182,
|
| 1276 |
+
"grad_norm": Infinity,
|
| 1277 |
+
"learning_rate": 9.468566648759864e-07,
|
| 1278 |
+
"loss": 2.9171165466308593,
|
| 1279 |
+
"step": 1630
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.14909090909090908,
|
| 1283 |
+
"grad_norm": Infinity,
|
| 1284 |
+
"learning_rate": 9.462141890618589e-07,
|
| 1285 |
+
"loss": 2.7405115127563477,
|
| 1286 |
+
"step": 1640
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.15,
|
| 1290 |
+
"grad_norm": 8.394250109558194e+18,
|
| 1291 |
+
"learning_rate": 9.455680736158879e-07,
|
| 1292 |
+
"loss": 2.944680404663086,
|
| 1293 |
+
"step": 1650
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.1509090909090909,
|
| 1297 |
+
"grad_norm": Infinity,
|
| 1298 |
+
"learning_rate": 9.449183238082383e-07,
|
| 1299 |
+
"loss": 3.1834575653076174,
|
| 1300 |
+
"step": 1660
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.15181818181818182,
|
| 1304 |
+
"grad_norm": Infinity,
|
| 1305 |
+
"learning_rate": 9.442649449387193e-07,
|
| 1306 |
+
"loss": 2.709722137451172,
|
| 1307 |
+
"step": 1670
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.15272727272727274,
|
| 1311 |
+
"grad_norm": Infinity,
|
| 1312 |
+
"learning_rate": 9.436079423367412e-07,
|
| 1313 |
+
"loss": 2.5861513137817385,
|
| 1314 |
+
"step": 1680
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.15363636363636363,
|
| 1318 |
+
"grad_norm": Infinity,
|
| 1319 |
+
"learning_rate": 9.429473213612722e-07,
|
| 1320 |
+
"loss": 3.1419662475585937,
|
| 1321 |
+
"step": 1690
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 0.15454545454545454,
|
| 1325 |
+
"grad_norm": Infinity,
|
| 1326 |
+
"learning_rate": 9.422830874007943e-07,
|
| 1327 |
+
"loss": 2.909741020202637,
|
| 1328 |
+
"step": 1700
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.15454545454545454,
|
| 1332 |
+
"eval_loss": 2.789142608642578,
|
| 1333 |
+
"eval_runtime": 25.3588,
|
| 1334 |
+
"eval_samples_per_second": 5.797,
|
| 1335 |
+
"eval_steps_per_second": 1.459,
|
| 1336 |
+
"step": 1700
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.15545454545454546,
|
| 1340 |
+
"grad_norm": Infinity,
|
| 1341 |
+
"learning_rate": 9.416152458732595e-07,
|
| 1342 |
+
"loss": 2.784795379638672,
|
| 1343 |
+
"step": 1710
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.15636363636363637,
|
| 1347 |
+
"grad_norm": Infinity,
|
| 1348 |
+
"learning_rate": 9.409438022260456e-07,
|
| 1349 |
+
"loss": 2.862109375,
|
| 1350 |
+
"step": 1720
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.1572727272727273,
|
| 1354 |
+
"grad_norm": Infinity,
|
| 1355 |
+
"learning_rate": 9.40268761935912e-07,
|
| 1356 |
+
"loss": 2.937261199951172,
|
| 1357 |
+
"step": 1730
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.15818181818181817,
|
| 1361 |
+
"grad_norm": 1.2604903555405447e+19,
|
| 1362 |
+
"learning_rate": 9.395901305089544e-07,
|
| 1363 |
+
"loss": 3.0458484649658204,
|
| 1364 |
+
"step": 1740
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.1590909090909091,
|
| 1368 |
+
"grad_norm": Infinity,
|
| 1369 |
+
"learning_rate": 9.389079134805609e-07,
|
| 1370 |
+
"loss": 2.753535270690918,
|
| 1371 |
+
"step": 1750
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.16,
|
| 1375 |
+
"grad_norm": Infinity,
|
| 1376 |
+
"learning_rate": 9.382221164153654e-07,
|
| 1377 |
+
"loss": 2.8955476760864256,
|
| 1378 |
+
"step": 1760
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.16090909090909092,
|
| 1382 |
+
"grad_norm": 7.74396099974031e+18,
|
| 1383 |
+
"learning_rate": 9.37532744907204e-07,
|
| 1384 |
+
"loss": 2.53248291015625,
|
| 1385 |
+
"step": 1770
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.1618181818181818,
|
| 1389 |
+
"grad_norm": 1.391111567262194e+19,
|
| 1390 |
+
"learning_rate": 9.368398045790678e-07,
|
| 1391 |
+
"loss": 2.7983562469482424,
|
| 1392 |
+
"step": 1780
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.16272727272727272,
|
| 1396 |
+
"grad_norm": Infinity,
|
| 1397 |
+
"learning_rate": 9.361433010830577e-07,
|
| 1398 |
+
"loss": 2.625954246520996,
|
| 1399 |
+
"step": 1790
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.16363636363636364,
|
| 1403 |
+
"grad_norm": 1.6281627557734908e+19,
|
| 1404 |
+
"learning_rate": 9.354432401003386e-07,
|
| 1405 |
+
"loss": 2.65600528717041,
|
| 1406 |
+
"step": 1800
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.16363636363636364,
|
| 1410 |
+
"eval_loss": 2.7892098426818848,
|
| 1411 |
+
"eval_runtime": 25.3434,
|
| 1412 |
+
"eval_samples_per_second": 5.8,
|
| 1413 |
+
"eval_steps_per_second": 1.46,
|
| 1414 |
+
"step": 1800
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 0.16454545454545455,
|
| 1418 |
+
"grad_norm": Infinity,
|
| 1419 |
+
"learning_rate": 9.347396273410924e-07,
|
| 1420 |
+
"loss": 2.858198547363281,
|
| 1421 |
+
"step": 1810
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 0.16545454545454547,
|
| 1425 |
+
"grad_norm": 1.5438372507974894e+19,
|
| 1426 |
+
"learning_rate": 9.34032468544472e-07,
|
| 1427 |
+
"loss": 2.5564403533935547,
|
| 1428 |
+
"step": 1820
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.16636363636363635,
|
| 1432 |
+
"grad_norm": Infinity,
|
| 1433 |
+
"learning_rate": 9.333217694785542e-07,
|
| 1434 |
+
"loss": 2.579339599609375,
|
| 1435 |
+
"step": 1830
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 0.16727272727272727,
|
| 1439 |
+
"grad_norm": 5.160486906466664e+18,
|
| 1440 |
+
"learning_rate": 9.326075359402924e-07,
|
| 1441 |
+
"loss": 2.883788299560547,
|
| 1442 |
+
"step": 1840
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.16818181818181818,
|
| 1446 |
+
"grad_norm": Infinity,
|
| 1447 |
+
"learning_rate": 9.318897737554698e-07,
|
| 1448 |
+
"loss": 2.588374900817871,
|
| 1449 |
+
"step": 1850
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.1690909090909091,
|
| 1453 |
+
"grad_norm": Infinity,
|
| 1454 |
+
"learning_rate": 9.31168488778652e-07,
|
| 1455 |
+
"loss": 2.779450035095215,
|
| 1456 |
+
"step": 1860
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 0.17,
|
| 1460 |
+
"grad_norm": Infinity,
|
| 1461 |
+
"learning_rate": 9.304436868931382e-07,
|
| 1462 |
+
"loss": 2.9528104782104494,
|
| 1463 |
+
"step": 1870
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.1709090909090909,
|
| 1467 |
+
"grad_norm": Infinity,
|
| 1468 |
+
"learning_rate": 9.297153740109147e-07,
|
| 1469 |
+
"loss": 3.0303653717041015,
|
| 1470 |
+
"step": 1880
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.17181818181818181,
|
| 1474 |
+
"grad_norm": 1.5527703429685182e+19,
|
| 1475 |
+
"learning_rate": 9.289835560726052e-07,
|
| 1476 |
+
"loss": 2.797162628173828,
|
| 1477 |
+
"step": 1890
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 0.17272727272727273,
|
| 1481 |
+
"grad_norm": 1.553810151114906e+19,
|
| 1482 |
+
"learning_rate": 9.28248239047424e-07,
|
| 1483 |
+
"loss": 2.6478057861328126,
|
| 1484 |
+
"step": 1900
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.17272727272727273,
|
| 1488 |
+
"eval_loss": 2.789252281188965,
|
| 1489 |
+
"eval_runtime": 25.3648,
|
| 1490 |
+
"eval_samples_per_second": 5.795,
|
| 1491 |
+
"eval_steps_per_second": 1.459,
|
| 1492 |
+
"step": 1900
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 0.17363636363636364,
|
| 1496 |
+
"grad_norm": Infinity,
|
| 1497 |
+
"learning_rate": 9.275094289331253e-07,
|
| 1498 |
+
"loss": 2.7918701171875,
|
| 1499 |
+
"step": 1910
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 0.17454545454545456,
|
| 1503 |
+
"grad_norm": Infinity,
|
| 1504 |
+
"learning_rate": 9.267671317559562e-07,
|
| 1505 |
+
"loss": 2.873759078979492,
|
| 1506 |
+
"step": 1920
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.17545454545454545,
|
| 1510 |
+
"grad_norm": 7.946095217390649e+18,
|
| 1511 |
+
"learning_rate": 9.260213535706062e-07,
|
| 1512 |
+
"loss": 3.0809797286987304,
|
| 1513 |
+
"step": 1930
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 0.17636363636363636,
|
| 1517 |
+
"grad_norm": Infinity,
|
| 1518 |
+
"learning_rate": 9.252721004601587e-07,
|
| 1519 |
+
"loss": 2.7745038986206056,
|
| 1520 |
+
"step": 1940
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 0.17727272727272728,
|
| 1524 |
+
"grad_norm": Infinity,
|
| 1525 |
+
"learning_rate": 9.245193785360406e-07,
|
| 1526 |
+
"loss": 2.998860740661621,
|
| 1527 |
+
"step": 1950
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 0.1781818181818182,
|
| 1531 |
+
"grad_norm": Infinity,
|
| 1532 |
+
"learning_rate": 9.23763193937973e-07,
|
| 1533 |
+
"loss": 2.7716712951660156,
|
| 1534 |
+
"step": 1960
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 0.17909090909090908,
|
| 1538 |
+
"grad_norm": Infinity,
|
| 1539 |
+
"learning_rate": 9.23003552833921e-07,
|
| 1540 |
+
"loss": 3.0993738174438477,
|
| 1541 |
+
"step": 1970
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 0.18,
|
| 1545 |
+
"grad_norm": 1.3247786904654447e+19,
|
| 1546 |
+
"learning_rate": 9.222404614200436e-07,
|
| 1547 |
+
"loss": 2.863362121582031,
|
| 1548 |
+
"step": 1980
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 0.1809090909090909,
|
| 1552 |
+
"grad_norm": Infinity,
|
| 1553 |
+
"learning_rate": 9.214739259206423e-07,
|
| 1554 |
+
"loss": 3.02794189453125,
|
| 1555 |
+
"step": 1990
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 0.18181818181818182,
|
| 1559 |
+
"grad_norm": Infinity,
|
| 1560 |
+
"learning_rate": 9.207039525881117e-07,
|
| 1561 |
+
"loss": 3.025308036804199,
|
| 1562 |
+
"step": 2000
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.18181818181818182,
|
| 1566 |
+
"eval_loss": 2.7894153594970703,
|
| 1567 |
+
"eval_runtime": 25.359,
|
| 1568 |
+
"eval_samples_per_second": 5.797,
|
| 1569 |
+
"eval_steps_per_second": 1.459,
|
| 1570 |
+
"step": 2000
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.18272727272727274,
|
| 1574 |
+
"grad_norm": Infinity,
|
| 1575 |
+
"learning_rate": 9.199305477028869e-07,
|
| 1576 |
+
"loss": 2.7843177795410154,
|
| 1577 |
+
"step": 2010
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.18363636363636363,
|
| 1581 |
+
"grad_norm": 5.96282857880132e+18,
|
| 1582 |
+
"learning_rate": 9.19153717573394e-07,
|
| 1583 |
+
"loss": 2.802363967895508,
|
| 1584 |
+
"step": 2020
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.18454545454545454,
|
| 1588 |
+
"grad_norm": Infinity,
|
| 1589 |
+
"learning_rate": 9.183734685359972e-07,
|
| 1590 |
+
"loss": 3.1789117813110352,
|
| 1591 |
+
"step": 2030
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.18545454545454546,
|
| 1595 |
+
"grad_norm": Infinity,
|
| 1596 |
+
"learning_rate": 9.175898069549476e-07,
|
| 1597 |
+
"loss": 2.3378616333007813,
|
| 1598 |
+
"step": 2040
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.18636363636363637,
|
| 1602 |
+
"grad_norm": Infinity,
|
| 1603 |
+
"learning_rate": 9.168027392223319e-07,
|
| 1604 |
+
"loss": 3.002253532409668,
|
| 1605 |
+
"step": 2050
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.18727272727272729,
|
| 1609 |
+
"grad_norm": 9.47185286863913e+18,
|
| 1610 |
+
"learning_rate": 9.160122717580194e-07,
|
| 1611 |
+
"loss": 2.89359073638916,
|
| 1612 |
+
"step": 2060
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.18818181818181817,
|
| 1616 |
+
"grad_norm": Infinity,
|
| 1617 |
+
"learning_rate": 9.152184110096097e-07,
|
| 1618 |
+
"loss": 2.685699462890625,
|
| 1619 |
+
"step": 2070
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.1890909090909091,
|
| 1623 |
+
"grad_norm": Infinity,
|
| 1624 |
+
"learning_rate": 9.144211634523808e-07,
|
| 1625 |
+
"loss": 3.139577293395996,
|
| 1626 |
+
"step": 2080
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.19,
|
| 1630 |
+
"grad_norm": 6.611679527410074e+18,
|
| 1631 |
+
"learning_rate": 9.13620535589236e-07,
|
| 1632 |
+
"loss": 2.7923168182373046,
|
| 1633 |
+
"step": 2090
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.19090909090909092,
|
| 1637 |
+
"grad_norm": Infinity,
|
| 1638 |
+
"learning_rate": 9.128165339506502e-07,
|
| 1639 |
+
"loss": 2.4308433532714844,
|
| 1640 |
+
"step": 2100
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.19090909090909092,
|
| 1644 |
+
"eval_loss": 2.789486885070801,
|
| 1645 |
+
"eval_runtime": 25.465,
|
| 1646 |
+
"eval_samples_per_second": 5.773,
|
| 1647 |
+
"eval_steps_per_second": 1.453,
|
| 1648 |
+
"step": 2100
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.1918181818181818,
|
| 1652 |
+
"grad_norm": Infinity,
|
| 1653 |
+
"learning_rate": 9.120091650946171e-07,
|
| 1654 |
+
"loss": 3.2387130737304686,
|
| 1655 |
+
"step": 2110
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.19272727272727272,
|
| 1659 |
+
"grad_norm": Infinity,
|
| 1660 |
+
"learning_rate": 9.111984356065966e-07,
|
| 1661 |
+
"loss": 3.4744667053222655,
|
| 1662 |
+
"step": 2120
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.19363636363636363,
|
| 1666 |
+
"grad_norm": 1.4776962387797869e+19,
|
| 1667 |
+
"learning_rate": 9.103843520994592e-07,
|
| 1668 |
+
"loss": 3.85797233581543,
|
| 1669 |
+
"step": 2130
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.19454545454545455,
|
| 1673 |
+
"grad_norm": Infinity,
|
| 1674 |
+
"learning_rate": 9.095669212134339e-07,
|
| 1675 |
+
"loss": 3.818370819091797,
|
| 1676 |
+
"step": 2140
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.19545454545454546,
|
| 1680 |
+
"grad_norm": Infinity,
|
| 1681 |
+
"learning_rate": 9.087461496160528e-07,
|
| 1682 |
+
"loss": 3.4372634887695312,
|
| 1683 |
+
"step": 2150
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.19636363636363635,
|
| 1687 |
+
"grad_norm": Infinity,
|
| 1688 |
+
"learning_rate": 9.079220440020972e-07,
|
| 1689 |
+
"loss": 2.8101428985595702,
|
| 1690 |
+
"step": 2160
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.19727272727272727,
|
| 1694 |
+
"grad_norm": Infinity,
|
| 1695 |
+
"learning_rate": 9.070946110935431e-07,
|
| 1696 |
+
"loss": 2.7843332290649414,
|
| 1697 |
+
"step": 2170
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.19818181818181818,
|
| 1701 |
+
"grad_norm": Infinity,
|
| 1702 |
+
"learning_rate": 9.062638576395061e-07,
|
| 1703 |
+
"loss": 2.402296447753906,
|
| 1704 |
+
"step": 2180
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.1990909090909091,
|
| 1708 |
+
"grad_norm": Infinity,
|
| 1709 |
+
"learning_rate": 9.054297904161867e-07,
|
| 1710 |
+
"loss": 2.8248491287231445,
|
| 1711 |
+
"step": 2190
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.2,
|
| 1715 |
+
"grad_norm": Infinity,
|
| 1716 |
+
"learning_rate": 9.045924162268144e-07,
|
| 1717 |
+
"loss": 2.7025005340576174,
|
| 1718 |
+
"step": 2200
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.2,
|
| 1722 |
+
"eval_loss": 2.7895781993865967,
|
| 1723 |
+
"eval_runtime": 25.4794,
|
| 1724 |
+
"eval_samples_per_second": 5.769,
|
| 1725 |
+
"eval_steps_per_second": 1.452,
|
| 1726 |
+
"step": 2200
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.2009090909090909,
|
| 1730 |
+
"grad_norm": 1.249766048976706e+19,
|
| 1731 |
+
"learning_rate": 9.037517419015928e-07,
|
| 1732 |
+
"loss": 2.798247528076172,
|
| 1733 |
+
"step": 2210
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.2018181818181818,
|
| 1737 |
+
"grad_norm": 1.637043731093363e+19,
|
| 1738 |
+
"learning_rate": 9.029077742976437e-07,
|
| 1739 |
+
"loss": 2.894302177429199,
|
| 1740 |
+
"step": 2220
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.20272727272727273,
|
| 1744 |
+
"grad_norm": Infinity,
|
| 1745 |
+
"learning_rate": 9.020605202989513e-07,
|
| 1746 |
+
"loss": 2.6775253295898436,
|
| 1747 |
+
"step": 2230
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.20363636363636364,
|
| 1751 |
+
"grad_norm": Infinity,
|
| 1752 |
+
"learning_rate": 9.012099868163056e-07,
|
| 1753 |
+
"loss": 2.829499053955078,
|
| 1754 |
+
"step": 2240
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.20454545454545456,
|
| 1758 |
+
"grad_norm": 1.7267504658780258e+19,
|
| 1759 |
+
"learning_rate": 9.003561807872469e-07,
|
| 1760 |
+
"loss": 2.526635932922363,
|
| 1761 |
+
"step": 2250
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.20545454545454545,
|
| 1765 |
+
"grad_norm": Infinity,
|
| 1766 |
+
"learning_rate": 8.994991091760076e-07,
|
| 1767 |
+
"loss": 2.897632598876953,
|
| 1768 |
+
"step": 2260
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.20636363636363636,
|
| 1772 |
+
"grad_norm": Infinity,
|
| 1773 |
+
"learning_rate": 8.986387789734576e-07,
|
| 1774 |
+
"loss": 2.66369514465332,
|
| 1775 |
+
"step": 2270
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.20727272727272728,
|
| 1779 |
+
"grad_norm": Infinity,
|
| 1780 |
+
"learning_rate": 8.977751971970456e-07,
|
| 1781 |
+
"loss": 2.954084587097168,
|
| 1782 |
+
"step": 2280
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.2081818181818182,
|
| 1786 |
+
"grad_norm": 4.468861107246727e+18,
|
| 1787 |
+
"learning_rate": 8.969083708907424e-07,
|
| 1788 |
+
"loss": 3.04129581451416,
|
| 1789 |
+
"step": 2290
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.20909090909090908,
|
| 1793 |
+
"grad_norm": Infinity,
|
| 1794 |
+
"learning_rate": 8.960383071249835e-07,
|
| 1795 |
+
"loss": 2.8931554794311523,
|
| 1796 |
+
"step": 2300
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.20909090909090908,
|
| 1800 |
+
"eval_loss": 2.78959321975708,
|
| 1801 |
+
"eval_runtime": 25.2525,
|
| 1802 |
+
"eval_samples_per_second": 5.821,
|
| 1803 |
+
"eval_steps_per_second": 1.465,
|
| 1804 |
+
"step": 2300
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.21,
|
| 1808 |
+
"grad_norm": Infinity,
|
| 1809 |
+
"learning_rate": 8.951650129966112e-07,
|
| 1810 |
+
"loss": 2.716269683837891,
|
| 1811 |
+
"step": 2310
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.2109090909090909,
|
| 1815 |
+
"grad_norm": Infinity,
|
| 1816 |
+
"learning_rate": 8.942884956288171e-07,
|
| 1817 |
+
"loss": 2.980978012084961,
|
| 1818 |
+
"step": 2320
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.21181818181818182,
|
| 1822 |
+
"grad_norm": 6.872592537169691e+18,
|
| 1823 |
+
"learning_rate": 8.934087621710835e-07,
|
| 1824 |
+
"loss": 2.609796905517578,
|
| 1825 |
+
"step": 2330
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.21272727272727274,
|
| 1829 |
+
"grad_norm": Infinity,
|
| 1830 |
+
"learning_rate": 8.925258197991258e-07,
|
| 1831 |
+
"loss": 3.0382633209228516,
|
| 1832 |
+
"step": 2340
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.21363636363636362,
|
| 1836 |
+
"grad_norm": Infinity,
|
| 1837 |
+
"learning_rate": 8.91639675714833e-07,
|
| 1838 |
+
"loss": 3.127687454223633,
|
| 1839 |
+
"step": 2350
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.21454545454545454,
|
| 1843 |
+
"grad_norm": Infinity,
|
| 1844 |
+
"learning_rate": 8.907503371462099e-07,
|
| 1845 |
+
"loss": 2.8488794326782227,
|
| 1846 |
+
"step": 2360
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.21545454545454545,
|
| 1850 |
+
"grad_norm": Infinity,
|
| 1851 |
+
"learning_rate": 8.898578113473176e-07,
|
| 1852 |
+
"loss": 2.7067138671875,
|
| 1853 |
+
"step": 2370
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.21636363636363637,
|
| 1857 |
+
"grad_norm": Infinity,
|
| 1858 |
+
"learning_rate": 8.889621055982144e-07,
|
| 1859 |
+
"loss": 2.789401626586914,
|
| 1860 |
+
"step": 2380
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.21727272727272728,
|
| 1864 |
+
"grad_norm": Infinity,
|
| 1865 |
+
"learning_rate": 8.880632272048962e-07,
|
| 1866 |
+
"loss": 3.3033519744873048,
|
| 1867 |
+
"step": 2390
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.21818181818181817,
|
| 1871 |
+
"grad_norm": Infinity,
|
| 1872 |
+
"learning_rate": 8.871611834992379e-07,
|
| 1873 |
+
"loss": 2.878365707397461,
|
| 1874 |
+
"step": 2400
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.21818181818181817,
|
| 1878 |
+
"eval_loss": 2.789731025695801,
|
| 1879 |
+
"eval_runtime": 25.3629,
|
| 1880 |
+
"eval_samples_per_second": 5.796,
|
| 1881 |
+
"eval_steps_per_second": 1.459,
|
| 1882 |
+
"step": 2400
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 0.2190909090909091,
|
| 1886 |
+
"grad_norm": Infinity,
|
| 1887 |
+
"learning_rate": 8.862559818389321e-07,
|
| 1888 |
+
"loss": 2.9361312866210936,
|
| 1889 |
+
"step": 2410
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 0.22,
|
| 1893 |
+
"grad_norm": 1.4403865107144638e+19,
|
| 1894 |
+
"learning_rate": 8.853476296074305e-07,
|
| 1895 |
+
"loss": 2.718800735473633,
|
| 1896 |
+
"step": 2420
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 0.22090909090909092,
|
| 1900 |
+
"grad_norm": Infinity,
|
| 1901 |
+
"learning_rate": 8.844361342138827e-07,
|
| 1902 |
+
"loss": 2.7465553283691406,
|
| 1903 |
+
"step": 2430
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 0.22181818181818183,
|
| 1907 |
+
"grad_norm": Infinity,
|
| 1908 |
+
"learning_rate": 8.835215030930761e-07,
|
| 1909 |
+
"loss": 2.5267559051513673,
|
| 1910 |
+
"step": 2440
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 0.22272727272727272,
|
| 1914 |
+
"grad_norm": Infinity,
|
| 1915 |
+
"learning_rate": 8.82603743705375e-07,
|
| 1916 |
+
"loss": 2.8523298263549806,
|
| 1917 |
+
"step": 2450
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 0.22363636363636363,
|
| 1921 |
+
"grad_norm": Infinity,
|
| 1922 |
+
"learning_rate": 8.81682863536661e-07,
|
| 1923 |
+
"loss": 2.890406608581543,
|
| 1924 |
+
"step": 2460
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 0.22454545454545455,
|
| 1928 |
+
"grad_norm": Infinity,
|
| 1929 |
+
"learning_rate": 8.807588700982699e-07,
|
| 1930 |
+
"loss": 3.1316598892211913,
|
| 1931 |
+
"step": 2470
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 0.22545454545454546,
|
| 1935 |
+
"grad_norm": Infinity,
|
| 1936 |
+
"learning_rate": 8.798317709269318e-07,
|
| 1937 |
+
"loss": 2.738657760620117,
|
| 1938 |
+
"step": 2480
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 0.22636363636363635,
|
| 1942 |
+
"grad_norm": Infinity,
|
| 1943 |
+
"learning_rate": 8.789015735847098e-07,
|
| 1944 |
+
"loss": 2.8348062515258787,
|
| 1945 |
+
"step": 2490
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 0.22727272727272727,
|
| 1949 |
+
"grad_norm": Infinity,
|
| 1950 |
+
"learning_rate": 8.779682856589367e-07,
|
| 1951 |
+
"loss": 3.0129791259765626,
|
| 1952 |
+
"step": 2500
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 0.22727272727272727,
|
| 1956 |
+
"eval_loss": 2.7899017333984375,
|
| 1957 |
+
"eval_runtime": 25.3362,
|
| 1958 |
+
"eval_samples_per_second": 5.802,
|
| 1959 |
+
"eval_steps_per_second": 1.46,
|
| 1960 |
+
"step": 2500
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 0.22818181818181818,
|
| 1964 |
+
"grad_norm": Infinity,
|
| 1965 |
+
"learning_rate": 8.770319147621557e-07,
|
| 1966 |
+
"loss": 2.794989013671875,
|
| 1967 |
+
"step": 2510
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 0.2290909090909091,
|
| 1971 |
+
"grad_norm": Infinity,
|
| 1972 |
+
"learning_rate": 8.760924685320557e-07,
|
| 1973 |
+
"loss": 2.7114553451538086,
|
| 1974 |
+
"step": 2520
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 0.23,
|
| 1978 |
+
"grad_norm": Infinity,
|
| 1979 |
+
"learning_rate": 8.751499546314105e-07,
|
| 1980 |
+
"loss": 2.5232223510742187,
|
| 1981 |
+
"step": 2530
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.2309090909090909,
|
| 1985 |
+
"grad_norm": Infinity,
|
| 1986 |
+
"learning_rate": 8.742043807480162e-07,
|
| 1987 |
+
"loss": 2.8785938262939452,
|
| 1988 |
+
"step": 2540
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 0.2318181818181818,
|
| 1992 |
+
"grad_norm": Infinity,
|
| 1993 |
+
"learning_rate": 8.732557545946278e-07,
|
| 1994 |
+
"loss": 2.575123596191406,
|
| 1995 |
+
"step": 2550
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.23272727272727273,
|
| 1999 |
+
"grad_norm": Infinity,
|
| 2000 |
+
"learning_rate": 8.723040839088969e-07,
|
| 2001 |
+
"loss": 2.522659683227539,
|
| 2002 |
+
"step": 2560
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 0.23363636363636364,
|
| 2006 |
+
"grad_norm": Infinity,
|
| 2007 |
+
"learning_rate": 8.713493764533088e-07,
|
| 2008 |
+
"loss": 2.8160863876342774,
|
| 2009 |
+
"step": 2570
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.23454545454545456,
|
| 2013 |
+
"grad_norm": 1.7132498099906806e+18,
|
| 2014 |
+
"learning_rate": 8.703916400151183e-07,
|
| 2015 |
+
"loss": 2.9550497055053713,
|
| 2016 |
+
"step": 2580
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.23545454545454544,
|
| 2020 |
+
"grad_norm": Infinity,
|
| 2021 |
+
"learning_rate": 8.694308824062867e-07,
|
| 2022 |
+
"loss": 2.701577568054199,
|
| 2023 |
+
"step": 2590
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.23636363636363636,
|
| 2027 |
+
"grad_norm": Infinity,
|
| 2028 |
+
"learning_rate": 8.684671114634183e-07,
|
| 2029 |
+
"loss": 2.701395606994629,
|
| 2030 |
+
"step": 2600
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 0.23636363636363636,
|
| 2034 |
+
"eval_loss": 2.7901124954223633,
|
| 2035 |
+
"eval_runtime": 25.2085,
|
| 2036 |
+
"eval_samples_per_second": 5.831,
|
| 2037 |
+
"eval_steps_per_second": 1.468,
|
| 2038 |
+
"step": 2600
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 0.23727272727272727,
|
| 2042 |
+
"grad_norm": Infinity,
|
| 2043 |
+
"learning_rate": 8.675003350476961e-07,
|
| 2044 |
+
"loss": 3.045684814453125,
|
| 2045 |
+
"step": 2610
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 0.2381818181818182,
|
| 2049 |
+
"grad_norm": 1.7417936541157687e+19,
|
| 2050 |
+
"learning_rate": 8.66530561044818e-07,
|
| 2051 |
+
"loss": 3.0101932525634765,
|
| 2052 |
+
"step": 2620
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 0.2390909090909091,
|
| 2056 |
+
"grad_norm": 1.790928409835497e+19,
|
| 2057 |
+
"learning_rate": 8.655577973649321e-07,
|
| 2058 |
+
"loss": 2.312086486816406,
|
| 2059 |
+
"step": 2630
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 0.24,
|
| 2063 |
+
"grad_norm": Infinity,
|
| 2064 |
+
"learning_rate": 8.645820519425722e-07,
|
| 2065 |
+
"loss": 3.022278594970703,
|
| 2066 |
+
"step": 2640
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 0.2409090909090909,
|
| 2070 |
+
"grad_norm": Infinity,
|
| 2071 |
+
"learning_rate": 8.636033327365937e-07,
|
| 2072 |
+
"loss": 2.964511489868164,
|
| 2073 |
+
"step": 2650
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 0.24181818181818182,
|
| 2077 |
+
"grad_norm": 6.040087962350256e+18,
|
| 2078 |
+
"learning_rate": 8.626216477301081e-07,
|
| 2079 |
+
"loss": 3.050629997253418,
|
| 2080 |
+
"step": 2660
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 0.24272727272727274,
|
| 2084 |
+
"grad_norm": Infinity,
|
| 2085 |
+
"learning_rate": 8.616370049304175e-07,
|
| 2086 |
+
"loss": 2.869434356689453,
|
| 2087 |
+
"step": 2670
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 0.24363636363636362,
|
| 2091 |
+
"grad_norm": 1.4910112146409914e+19,
|
| 2092 |
+
"learning_rate": 8.606494123689508e-07,
|
| 2093 |
+
"loss": 2.7562395095825196,
|
| 2094 |
+
"step": 2680
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 0.24454545454545454,
|
| 2098 |
+
"grad_norm": Infinity,
|
| 2099 |
+
"learning_rate": 8.596588781011963e-07,
|
| 2100 |
+
"loss": 2.5833467483520507,
|
| 2101 |
+
"step": 2690
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 0.24545454545454545,
|
| 2105 |
+
"grad_norm": 1.4346714691756098e+19,
|
| 2106 |
+
"learning_rate": 8.586654102066374e-07,
|
| 2107 |
+
"loss": 2.8177486419677735,
|
| 2108 |
+
"step": 2700
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 0.24545454545454545,
|
| 2112 |
+
"eval_loss": 2.790109157562256,
|
| 2113 |
+
"eval_runtime": 25.304,
|
| 2114 |
+
"eval_samples_per_second": 5.809,
|
| 2115 |
+
"eval_steps_per_second": 1.462,
|
| 2116 |
+
"step": 2700
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.24636363636363637,
|
| 2120 |
+
"grad_norm": Infinity,
|
| 2121 |
+
"learning_rate": 8.576690167886859e-07,
|
| 2122 |
+
"loss": 2.540431785583496,
|
| 2123 |
+
"step": 2710
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.24727272727272728,
|
| 2127 |
+
"grad_norm": 2.7262607964252406e+18,
|
| 2128 |
+
"learning_rate": 8.566697059746166e-07,
|
| 2129 |
+
"loss": 2.6125823974609377,
|
| 2130 |
+
"step": 2720
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.24818181818181817,
|
| 2134 |
+
"grad_norm": Infinity,
|
| 2135 |
+
"learning_rate": 8.556674859155e-07,
|
| 2136 |
+
"loss": 2.758362579345703,
|
| 2137 |
+
"step": 2730
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.24909090909090909,
|
| 2141 |
+
"grad_norm": Infinity,
|
| 2142 |
+
"learning_rate": 8.546623647861369e-07,
|
| 2143 |
+
"loss": 3.0073740005493166,
|
| 2144 |
+
"step": 2740
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.25,
|
| 2148 |
+
"grad_norm": 7.052120795752956e+18,
|
| 2149 |
+
"learning_rate": 8.536543507849911e-07,
|
| 2150 |
+
"loss": 2.758180618286133,
|
| 2151 |
+
"step": 2750
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.2509090909090909,
|
| 2155 |
+
"grad_norm": Infinity,
|
| 2156 |
+
"learning_rate": 8.526434521341226e-07,
|
| 2157 |
+
"loss": 2.963096046447754,
|
| 2158 |
+
"step": 2760
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.25181818181818183,
|
| 2162 |
+
"grad_norm": 1.3421638384703504e+19,
|
| 2163 |
+
"learning_rate": 8.516296770791207e-07,
|
| 2164 |
+
"loss": 2.8563003540039062,
|
| 2165 |
+
"step": 2770
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.25272727272727274,
|
| 2169 |
+
"grad_norm": Infinity,
|
| 2170 |
+
"learning_rate": 8.506130338890365e-07,
|
| 2171 |
+
"loss": 3.0656408309936523,
|
| 2172 |
+
"step": 2780
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.25363636363636366,
|
| 2176 |
+
"grad_norm": Infinity,
|
| 2177 |
+
"learning_rate": 8.495935308563158e-07,
|
| 2178 |
+
"loss": 2.983555221557617,
|
| 2179 |
+
"step": 2790
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.2545454545454545,
|
| 2183 |
+
"grad_norm": Infinity,
|
| 2184 |
+
"learning_rate": 8.485711762967312e-07,
|
| 2185 |
+
"loss": 2.951832580566406,
|
| 2186 |
+
"step": 2800
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.2545454545454545,
|
| 2190 |
+
"eval_loss": 2.790069818496704,
|
| 2191 |
+
"eval_runtime": 25.4101,
|
| 2192 |
+
"eval_samples_per_second": 5.785,
|
| 2193 |
+
"eval_steps_per_second": 1.456,
|
| 2194 |
+
"step": 2800
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.25545454545454543,
|
| 2198 |
+
"grad_norm": Infinity,
|
| 2199 |
+
"learning_rate": 8.47545978549314e-07,
|
| 2200 |
+
"loss": 2.5605377197265624,
|
| 2201 |
+
"step": 2810
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.25636363636363635,
|
| 2205 |
+
"grad_norm": Infinity,
|
| 2206 |
+
"learning_rate": 8.46517945976287e-07,
|
| 2207 |
+
"loss": 2.6139829635620115,
|
| 2208 |
+
"step": 2820
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.25727272727272726,
|
| 2212 |
+
"grad_norm": 1.3888741710508327e+19,
|
| 2213 |
+
"learning_rate": 8.454870869629955e-07,
|
| 2214 |
+
"loss": 2.8513839721679686,
|
| 2215 |
+
"step": 2830
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.2581818181818182,
|
| 2219 |
+
"grad_norm": Infinity,
|
| 2220 |
+
"learning_rate": 8.444534099178393e-07,
|
| 2221 |
+
"loss": 2.7712711334228515,
|
| 2222 |
+
"step": 2840
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.2590909090909091,
|
| 2226 |
+
"grad_norm": Infinity,
|
| 2227 |
+
"learning_rate": 8.434169232722041e-07,
|
| 2228 |
+
"loss": 2.187318801879883,
|
| 2229 |
+
"step": 2850
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.26,
|
| 2233 |
+
"grad_norm": Infinity,
|
| 2234 |
+
"learning_rate": 8.423776354803925e-07,
|
| 2235 |
+
"loss": 3.0384477615356444,
|
| 2236 |
+
"step": 2860
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.2609090909090909,
|
| 2240 |
+
"grad_norm": Infinity,
|
| 2241 |
+
"learning_rate": 8.41335555019555e-07,
|
| 2242 |
+
"loss": 2.9354883193969727,
|
| 2243 |
+
"step": 2870
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.26181818181818184,
|
| 2247 |
+
"grad_norm": Infinity,
|
| 2248 |
+
"learning_rate": 8.402906903896216e-07,
|
| 2249 |
+
"loss": 3.0859096527099608,
|
| 2250 |
+
"step": 2880
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.26272727272727275,
|
| 2254 |
+
"grad_norm": Infinity,
|
| 2255 |
+
"learning_rate": 8.392430501132316e-07,
|
| 2256 |
+
"loss": 3.006024169921875,
|
| 2257 |
+
"step": 2890
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.2636363636363636,
|
| 2261 |
+
"grad_norm": Infinity,
|
| 2262 |
+
"learning_rate": 8.381926427356642e-07,
|
| 2263 |
+
"loss": 2.832804298400879,
|
| 2264 |
+
"step": 2900
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.2636363636363636,
|
| 2268 |
+
"eval_loss": 2.7902262210845947,
|
| 2269 |
+
"eval_runtime": 25.2761,
|
| 2270 |
+
"eval_samples_per_second": 5.816,
|
| 2271 |
+
"eval_steps_per_second": 1.464,
|
| 2272 |
+
"step": 2900
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 0.26454545454545453,
|
| 2276 |
+
"grad_norm": Infinity,
|
| 2277 |
+
"learning_rate": 8.37139476824769e-07,
|
| 2278 |
+
"loss": 2.8630512237548826,
|
| 2279 |
+
"step": 2910
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 0.26545454545454544,
|
| 2283 |
+
"grad_norm": Infinity,
|
| 2284 |
+
"learning_rate": 8.360835609708967e-07,
|
| 2285 |
+
"loss": 2.8531957626342774,
|
| 2286 |
+
"step": 2920
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 0.26636363636363636,
|
| 2290 |
+
"grad_norm": Infinity,
|
| 2291 |
+
"learning_rate": 8.35024903786828e-07,
|
| 2292 |
+
"loss": 2.766724967956543,
|
| 2293 |
+
"step": 2930
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.2672727272727273,
|
| 2297 |
+
"grad_norm": Infinity,
|
| 2298 |
+
"learning_rate": 8.339635139077035e-07,
|
| 2299 |
+
"loss": 3.115190887451172,
|
| 2300 |
+
"step": 2940
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 0.2681818181818182,
|
| 2304 |
+
"grad_norm": 1.7490984795172241e+19,
|
| 2305 |
+
"learning_rate": 8.32899399990954e-07,
|
| 2306 |
+
"loss": 3.0780994415283205,
|
| 2307 |
+
"step": 2950
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.2690909090909091,
|
| 2311 |
+
"grad_norm": Infinity,
|
| 2312 |
+
"learning_rate": 8.318325707162293e-07,
|
| 2313 |
+
"loss": 2.6672946929931642,
|
| 2314 |
+
"step": 2960
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 0.27,
|
| 2318 |
+
"grad_norm": Infinity,
|
| 2319 |
+
"learning_rate": 8.307630347853273e-07,
|
| 2320 |
+
"loss": 2.781101417541504,
|
| 2321 |
+
"step": 2970
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.27090909090909093,
|
| 2325 |
+
"grad_norm": Infinity,
|
| 2326 |
+
"learning_rate": 8.296908009221242e-07,
|
| 2327 |
+
"loss": 2.4955059051513673,
|
| 2328 |
+
"step": 2980
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 0.2718181818181818,
|
| 2332 |
+
"grad_norm": Infinity,
|
| 2333 |
+
"learning_rate": 8.286158778725011e-07,
|
| 2334 |
+
"loss": 2.4922977447509767,
|
| 2335 |
+
"step": 2990
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.2727272727272727,
|
| 2339 |
+
"grad_norm": Infinity,
|
| 2340 |
+
"learning_rate": 8.275382744042747e-07,
|
| 2341 |
+
"loss": 2.6550668716430663,
|
| 2342 |
+
"step": 3000
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.2727272727272727,
|
| 2346 |
+
"eval_loss": 2.7902653217315674,
|
| 2347 |
+
"eval_runtime": 25.2219,
|
| 2348 |
+
"eval_samples_per_second": 5.828,
|
| 2349 |
+
"eval_steps_per_second": 1.467,
|
| 2350 |
+
"step": 3000
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 0.2736363636363636,
|
| 2354 |
+
"grad_norm": 1.523740707118488e+19,
|
| 2355 |
+
"learning_rate": 8.26457999307125e-07,
|
| 2356 |
+
"loss": 2.947774887084961,
|
| 2357 |
+
"step": 3010
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.27454545454545454,
|
| 2361 |
+
"grad_norm": Infinity,
|
| 2362 |
+
"learning_rate": 8.253750613925234e-07,
|
| 2363 |
+
"loss": 2.497574806213379,
|
| 2364 |
+
"step": 3020
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.27545454545454545,
|
| 2368 |
+
"grad_norm": Infinity,
|
| 2369 |
+
"learning_rate": 8.242894694936614e-07,
|
| 2370 |
+
"loss": 2.8819416046142576,
|
| 2371 |
+
"step": 3030
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.27636363636363637,
|
| 2375 |
+
"grad_norm": Infinity,
|
| 2376 |
+
"learning_rate": 8.23201232465378e-07,
|
| 2377 |
+
"loss": 2.79040641784668,
|
| 2378 |
+
"step": 3040
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.2772727272727273,
|
| 2382 |
+
"grad_norm": Infinity,
|
| 2383 |
+
"learning_rate": 8.221103591840881e-07,
|
| 2384 |
+
"loss": 2.8814857482910154,
|
| 2385 |
+
"step": 3050
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.2781818181818182,
|
| 2389 |
+
"grad_norm": 7.556027524518511e+18,
|
| 2390 |
+
"learning_rate": 8.210168585477091e-07,
|
| 2391 |
+
"loss": 2.807586669921875,
|
| 2392 |
+
"step": 3060
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.2790909090909091,
|
| 2396 |
+
"grad_norm": Infinity,
|
| 2397 |
+
"learning_rate": 8.199207394755891e-07,
|
| 2398 |
+
"loss": 2.8781753540039063,
|
| 2399 |
+
"step": 3070
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.28,
|
| 2403 |
+
"grad_norm": Infinity,
|
| 2404 |
+
"learning_rate": 8.188220109084346e-07,
|
| 2405 |
+
"loss": 2.8258827209472654,
|
| 2406 |
+
"step": 3080
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.2809090909090909,
|
| 2410 |
+
"grad_norm": 7.713179072209093e+18,
|
| 2411 |
+
"learning_rate": 8.17720681808236e-07,
|
| 2412 |
+
"loss": 2.972319221496582,
|
| 2413 |
+
"step": 3090
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.2818181818181818,
|
| 2417 |
+
"grad_norm": Infinity,
|
| 2418 |
+
"learning_rate": 8.16616761158196e-07,
|
| 2419 |
+
"loss": 2.79774169921875,
|
| 2420 |
+
"step": 3100
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.2818181818181818,
|
| 2424 |
+
"eval_loss": 2.7904176712036133,
|
| 2425 |
+
"eval_runtime": 25.3955,
|
| 2426 |
+
"eval_samples_per_second": 5.788,
|
| 2427 |
+
"eval_steps_per_second": 1.457,
|
| 2428 |
+
"step": 3100
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 0.2827272727272727,
|
| 2432 |
+
"grad_norm": 6.054117180964864e+18,
|
| 2433 |
+
"learning_rate": 8.155102579626558e-07,
|
| 2434 |
+
"loss": 2.4805845260620116,
|
| 2435 |
+
"step": 3110
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 0.28363636363636363,
|
| 2439 |
+
"grad_norm": Infinity,
|
| 2440 |
+
"learning_rate": 8.144011812470215e-07,
|
| 2441 |
+
"loss": 2.973079299926758,
|
| 2442 |
+
"step": 3120
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 0.28454545454545455,
|
| 2446 |
+
"grad_norm": Infinity,
|
| 2447 |
+
"learning_rate": 8.132895400576904e-07,
|
| 2448 |
+
"loss": 2.4612403869628907,
|
| 2449 |
+
"step": 3130
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 0.28545454545454546,
|
| 2453 |
+
"grad_norm": Infinity,
|
| 2454 |
+
"learning_rate": 8.121753434619778e-07,
|
| 2455 |
+
"loss": 2.7316892623901365,
|
| 2456 |
+
"step": 3140
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 0.2863636363636364,
|
| 2460 |
+
"grad_norm": 1.4829298041768378e+19,
|
| 2461 |
+
"learning_rate": 8.110586005480426e-07,
|
| 2462 |
+
"loss": 2.7744102478027344,
|
| 2463 |
+
"step": 3150
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 0.2872727272727273,
|
| 2467 |
+
"grad_norm": Infinity,
|
| 2468 |
+
"learning_rate": 8.09939320424813e-07,
|
| 2469 |
+
"loss": 3.0601181030273437,
|
| 2470 |
+
"step": 3160
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 0.2881818181818182,
|
| 2474 |
+
"grad_norm": Infinity,
|
| 2475 |
+
"learning_rate": 8.088175122219127e-07,
|
| 2476 |
+
"loss": 3.284495544433594,
|
| 2477 |
+
"step": 3170
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 0.28909090909090907,
|
| 2481 |
+
"grad_norm": Infinity,
|
| 2482 |
+
"learning_rate": 8.076931850895858e-07,
|
| 2483 |
+
"loss": 3.0675275802612303,
|
| 2484 |
+
"step": 3180
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 0.29,
|
| 2488 |
+
"grad_norm": Infinity,
|
| 2489 |
+
"learning_rate": 8.065663481986229e-07,
|
| 2490 |
+
"loss": 2.9712600708007812,
|
| 2491 |
+
"step": 3190
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 0.2909090909090909,
|
| 2495 |
+
"grad_norm": Infinity,
|
| 2496 |
+
"learning_rate": 8.054370107402858e-07,
|
| 2497 |
+
"loss": 3.011086654663086,
|
| 2498 |
+
"step": 3200
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 0.2909090909090909,
|
| 2502 |
+
"eval_loss": 2.7904467582702637,
|
| 2503 |
+
"eval_runtime": 25.4518,
|
| 2504 |
+
"eval_samples_per_second": 5.776,
|
| 2505 |
+
"eval_steps_per_second": 1.454,
|
| 2506 |
+
"step": 3200
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 0.2918181818181818,
|
| 2510 |
+
"grad_norm": Infinity,
|
| 2511 |
+
"learning_rate": 8.043051819262327e-07,
|
| 2512 |
+
"loss": 2.8368200302124023,
|
| 2513 |
+
"step": 3210
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 0.2927272727272727,
|
| 2517 |
+
"grad_norm": Infinity,
|
| 2518 |
+
"learning_rate": 8.031708709884429e-07,
|
| 2519 |
+
"loss": 3.306093215942383,
|
| 2520 |
+
"step": 3220
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 0.29363636363636364,
|
| 2524 |
+
"grad_norm": 6.184249329914479e+18,
|
| 2525 |
+
"learning_rate": 8.020340871791417e-07,
|
| 2526 |
+
"loss": 2.9534244537353516,
|
| 2527 |
+
"step": 3230
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 0.29454545454545455,
|
| 2531 |
+
"grad_norm": 1.056424515862528e+19,
|
| 2532 |
+
"learning_rate": 8.008948397707252e-07,
|
| 2533 |
+
"loss": 2.3959999084472656,
|
| 2534 |
+
"step": 3240
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 0.29545454545454547,
|
| 2538 |
+
"grad_norm": Infinity,
|
| 2539 |
+
"learning_rate": 7.997531380556833e-07,
|
| 2540 |
+
"loss": 2.98052921295166,
|
| 2541 |
+
"step": 3250
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 0.2963636363636364,
|
| 2545 |
+
"grad_norm": Infinity,
|
| 2546 |
+
"learning_rate": 7.986089913465261e-07,
|
| 2547 |
+
"loss": 2.84487361907959,
|
| 2548 |
+
"step": 3260
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 0.2972727272727273,
|
| 2552 |
+
"grad_norm": Infinity,
|
| 2553 |
+
"learning_rate": 7.974624089757064e-07,
|
| 2554 |
+
"loss": 2.6605175018310545,
|
| 2555 |
+
"step": 3270
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 0.29818181818181816,
|
| 2559 |
+
"grad_norm": 7.253827303380091e+18,
|
| 2560 |
+
"learning_rate": 7.963134002955431e-07,
|
| 2561 |
+
"loss": 2.5641525268554686,
|
| 2562 |
+
"step": 3280
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 0.2990909090909091,
|
| 2566 |
+
"grad_norm": Infinity,
|
| 2567 |
+
"learning_rate": 7.951619746781472e-07,
|
| 2568 |
+
"loss": 2.7621686935424803,
|
| 2569 |
+
"step": 3290
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 0.3,
|
| 2573 |
+
"grad_norm": Infinity,
|
| 2574 |
+
"learning_rate": 7.940081415153428e-07,
|
| 2575 |
+
"loss": 2.810822296142578,
|
| 2576 |
+
"step": 3300
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 0.3,
|
| 2580 |
+
"eval_loss": 2.7905690670013428,
|
| 2581 |
+
"eval_runtime": 25.3785,
|
| 2582 |
+
"eval_samples_per_second": 5.792,
|
| 2583 |
+
"eval_steps_per_second": 1.458,
|
| 2584 |
+
"step": 3300
|
| 2585 |
+
},
|
| 2586 |
+
{
|
| 2587 |
+
"epoch": 0.3009090909090909,
|
| 2588 |
+
"grad_norm": Infinity,
|
| 2589 |
+
"learning_rate": 7.928519102185922e-07,
|
| 2590 |
+
"loss": 2.5717878341674805,
|
| 2591 |
+
"step": 3310
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 0.3018181818181818,
|
| 2595 |
+
"grad_norm": Infinity,
|
| 2596 |
+
"learning_rate": 7.91693290218918e-07,
|
| 2597 |
+
"loss": 2.659162902832031,
|
| 2598 |
+
"step": 3320
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 0.30272727272727273,
|
| 2602 |
+
"grad_norm": Infinity,
|
| 2603 |
+
"learning_rate": 7.905322909668272e-07,
|
| 2604 |
+
"loss": 2.7497512817382814,
|
| 2605 |
+
"step": 3330
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 0.30363636363636365,
|
| 2609 |
+
"grad_norm": Infinity,
|
| 2610 |
+
"learning_rate": 7.893689219322336e-07,
|
| 2611 |
+
"loss": 3.2085891723632813,
|
| 2612 |
+
"step": 3340
|
| 2613 |
+
},
|
| 2614 |
+
{
|
| 2615 |
+
"epoch": 0.30454545454545456,
|
| 2616 |
+
"grad_norm": Infinity,
|
| 2617 |
+
"learning_rate": 7.882031926043803e-07,
|
| 2618 |
+
"loss": 2.8847372055053713,
|
| 2619 |
+
"step": 3350
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 0.3054545454545455,
|
| 2623 |
+
"grad_norm": 7.063715695623668e+18,
|
| 2624 |
+
"learning_rate": 7.870351124917629e-07,
|
| 2625 |
+
"loss": 2.7339805603027343,
|
| 2626 |
+
"step": 3360
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 0.30636363636363634,
|
| 2630 |
+
"grad_norm": Infinity,
|
| 2631 |
+
"learning_rate": 7.858646911220512e-07,
|
| 2632 |
+
"loss": 2.932914924621582,
|
| 2633 |
+
"step": 3370
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 0.30727272727272725,
|
| 2637 |
+
"grad_norm": Infinity,
|
| 2638 |
+
"learning_rate": 7.846919380420125e-07,
|
| 2639 |
+
"loss": 3.179635238647461,
|
| 2640 |
+
"step": 3380
|
| 2641 |
+
},
|
| 2642 |
+
{
|
| 2643 |
+
"epoch": 0.30818181818181817,
|
| 2644 |
+
"grad_norm": Infinity,
|
| 2645 |
+
"learning_rate": 7.835168628174325e-07,
|
| 2646 |
+
"loss": 2.679301071166992,
|
| 2647 |
+
"step": 3390
|
| 2648 |
+
},
|
| 2649 |
+
{
|
| 2650 |
+
"epoch": 0.3090909090909091,
|
| 2651 |
+
"grad_norm": Infinity,
|
| 2652 |
+
"learning_rate": 7.823394750330386e-07,
|
| 2653 |
+
"loss": 2.8976104736328123,
|
| 2654 |
+
"step": 3400
|
| 2655 |
+
},
|
| 2656 |
+
{
|
| 2657 |
+
"epoch": 0.3090909090909091,
|
| 2658 |
+
"eval_loss": 2.7906525135040283,
|
| 2659 |
+
"eval_runtime": 25.3974,
|
| 2660 |
+
"eval_samples_per_second": 5.788,
|
| 2661 |
+
"eval_steps_per_second": 1.457,
|
| 2662 |
+
"step": 3400
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.31,
|
| 2666 |
+
"grad_norm": Infinity,
|
| 2667 |
+
"learning_rate": 7.811597842924203e-07,
|
| 2668 |
+
"loss": 3.08035888671875,
|
| 2669 |
+
"step": 3410
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.3109090909090909,
|
| 2673 |
+
"grad_norm": Infinity,
|
| 2674 |
+
"learning_rate": 7.799778002179523e-07,
|
| 2675 |
+
"loss": 2.6392980575561524,
|
| 2676 |
+
"step": 3420
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.3118181818181818,
|
| 2680 |
+
"grad_norm": Infinity,
|
| 2681 |
+
"learning_rate": 7.787935324507149e-07,
|
| 2682 |
+
"loss": 2.597859001159668,
|
| 2683 |
+
"step": 3430
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.31272727272727274,
|
| 2687 |
+
"grad_norm": Infinity,
|
| 2688 |
+
"learning_rate": 7.776069906504159e-07,
|
| 2689 |
+
"loss": 2.9615325927734375,
|
| 2690 |
+
"step": 3440
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.31363636363636366,
|
| 2694 |
+
"grad_norm": Infinity,
|
| 2695 |
+
"learning_rate": 7.764181844953116e-07,
|
| 2696 |
+
"loss": 2.8327489852905274,
|
| 2697 |
+
"step": 3450
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.3145454545454546,
|
| 2701 |
+
"grad_norm": 9.005495011717939e+18,
|
| 2702 |
+
"learning_rate": 7.752271236821282e-07,
|
| 2703 |
+
"loss": 2.8600826263427734,
|
| 2704 |
+
"step": 3460
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.31545454545454543,
|
| 2708 |
+
"grad_norm": 4.708173111567057e+18,
|
| 2709 |
+
"learning_rate": 7.74033817925982e-07,
|
| 2710 |
+
"loss": 2.58587646484375,
|
| 2711 |
+
"step": 3470
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.31636363636363635,
|
| 2715 |
+
"grad_norm": Infinity,
|
| 2716 |
+
"learning_rate": 7.728382769603011e-07,
|
| 2717 |
+
"loss": 2.7968507766723634,
|
| 2718 |
+
"step": 3480
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.31727272727272726,
|
| 2722 |
+
"grad_norm": Infinity,
|
| 2723 |
+
"learning_rate": 7.716405105367452e-07,
|
| 2724 |
+
"loss": 2.4269617080688475,
|
| 2725 |
+
"step": 3490
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.3181818181818182,
|
| 2729 |
+
"grad_norm": Infinity,
|
| 2730 |
+
"learning_rate": 7.704405284251267e-07,
|
| 2731 |
+
"loss": 2.9382980346679686,
|
| 2732 |
+
"step": 3500
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.3181818181818182,
|
| 2736 |
+
"eval_loss": 2.790801763534546,
|
| 2737 |
+
"eval_runtime": 25.2758,
|
| 2738 |
+
"eval_samples_per_second": 5.816,
|
| 2739 |
+
"eval_steps_per_second": 1.464,
|
| 2740 |
+
"step": 3500
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.3190909090909091,
|
| 2744 |
+
"grad_norm": Infinity,
|
| 2745 |
+
"learning_rate": 7.6923834041333e-07,
|
| 2746 |
+
"loss": 3.010089111328125,
|
| 2747 |
+
"step": 3510
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.32,
|
| 2751 |
+
"grad_norm": 1.7403037059089695e+19,
|
| 2752 |
+
"learning_rate": 7.680339563072333e-07,
|
| 2753 |
+
"loss": 3.1997365951538086,
|
| 2754 |
+
"step": 3520
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.3209090909090909,
|
| 2758 |
+
"grad_norm": Infinity,
|
| 2759 |
+
"learning_rate": 7.668273859306269e-07,
|
| 2760 |
+
"loss": 2.353929138183594,
|
| 2761 |
+
"step": 3530
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.32181818181818184,
|
| 2765 |
+
"grad_norm": Infinity,
|
| 2766 |
+
"learning_rate": 7.656186391251341e-07,
|
| 2767 |
+
"loss": 2.7316473007202147,
|
| 2768 |
+
"step": 3540
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.32272727272727275,
|
| 2772 |
+
"grad_norm": Infinity,
|
| 2773 |
+
"learning_rate": 7.644077257501308e-07,
|
| 2774 |
+
"loss": 2.843223762512207,
|
| 2775 |
+
"step": 3550
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.3236363636363636,
|
| 2779 |
+
"grad_norm": Infinity,
|
| 2780 |
+
"learning_rate": 7.631946556826647e-07,
|
| 2781 |
+
"loss": 2.877858543395996,
|
| 2782 |
+
"step": 3560
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.3245454545454545,
|
| 2786 |
+
"grad_norm": Infinity,
|
| 2787 |
+
"learning_rate": 7.619794388173751e-07,
|
| 2788 |
+
"loss": 2.1713531494140623,
|
| 2789 |
+
"step": 3570
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.32545454545454544,
|
| 2793 |
+
"grad_norm": Infinity,
|
| 2794 |
+
"learning_rate": 7.60762085066412e-07,
|
| 2795 |
+
"loss": 2.820762825012207,
|
| 2796 |
+
"step": 3580
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.32636363636363636,
|
| 2800 |
+
"grad_norm": 1.3926951938596798e+19,
|
| 2801 |
+
"learning_rate": 7.595426043593556e-07,
|
| 2802 |
+
"loss": 2.7209442138671873,
|
| 2803 |
+
"step": 3590
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.32727272727272727,
|
| 2807 |
+
"grad_norm": 9.029549577354609e+18,
|
| 2808 |
+
"learning_rate": 7.583210066431346e-07,
|
| 2809 |
+
"loss": 2.9243946075439453,
|
| 2810 |
+
"step": 3600
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.32727272727272727,
|
| 2814 |
+
"eval_loss": 2.790870189666748,
|
| 2815 |
+
"eval_runtime": 25.1317,
|
| 2816 |
+
"eval_samples_per_second": 5.849,
|
| 2817 |
+
"eval_steps_per_second": 1.472,
|
| 2818 |
+
"step": 3600
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 0.3281818181818182,
|
| 2822 |
+
"grad_norm": 7.039327015891108e+17,
|
| 2823 |
+
"learning_rate": 7.570973018819458e-07,
|
| 2824 |
+
"loss": 2.6024728775024415,
|
| 2825 |
+
"step": 3610
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.3290909090909091,
|
| 2829 |
+
"grad_norm": Infinity,
|
| 2830 |
+
"learning_rate": 7.558715000571725e-07,
|
| 2831 |
+
"loss": 2.9454456329345704,
|
| 2832 |
+
"step": 3620
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.33,
|
| 2836 |
+
"grad_norm": Infinity,
|
| 2837 |
+
"learning_rate": 7.546436111673027e-07,
|
| 2838 |
+
"loss": 2.9420135498046873,
|
| 2839 |
+
"step": 3630
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.33090909090909093,
|
| 2843 |
+
"grad_norm": Infinity,
|
| 2844 |
+
"learning_rate": 7.534136452278486e-07,
|
| 2845 |
+
"loss": 2.928874206542969,
|
| 2846 |
+
"step": 3640
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.33181818181818185,
|
| 2850 |
+
"grad_norm": Infinity,
|
| 2851 |
+
"learning_rate": 7.521816122712637e-07,
|
| 2852 |
+
"loss": 2.8279190063476562,
|
| 2853 |
+
"step": 3650
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.3327272727272727,
|
| 2857 |
+
"grad_norm": 1.0277020835612983e+19,
|
| 2858 |
+
"learning_rate": 7.509475223468618e-07,
|
| 2859 |
+
"loss": 2.930580902099609,
|
| 2860 |
+
"step": 3660
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.3336363636363636,
|
| 2864 |
+
"grad_norm": Infinity,
|
| 2865 |
+
"learning_rate": 7.497113855207347e-07,
|
| 2866 |
+
"loss": 2.815085220336914,
|
| 2867 |
+
"step": 3670
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.33454545454545453,
|
| 2871 |
+
"grad_norm": Infinity,
|
| 2872 |
+
"learning_rate": 7.4847321187567e-07,
|
| 2873 |
+
"loss": 2.9397621154785156,
|
| 2874 |
+
"step": 3680
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.33545454545454545,
|
| 2878 |
+
"grad_norm": 1.3450457683979141e+19,
|
| 2879 |
+
"learning_rate": 7.472330115110696e-07,
|
| 2880 |
+
"loss": 2.5939712524414062,
|
| 2881 |
+
"step": 3690
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.33636363636363636,
|
| 2885 |
+
"grad_norm": 1.2687692382930469e+19,
|
| 2886 |
+
"learning_rate": 7.459907945428657e-07,
|
| 2887 |
+
"loss": 3.0160362243652346,
|
| 2888 |
+
"step": 3700
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 0.33636363636363636,
|
| 2892 |
+
"eval_loss": 2.7909200191497803,
|
| 2893 |
+
"eval_runtime": 25.1818,
|
| 2894 |
+
"eval_samples_per_second": 5.838,
|
| 2895 |
+
"eval_steps_per_second": 1.469,
|
| 2896 |
+
"step": 3700
|
| 2897 |
+
},
|
| 2898 |
+
{
|
| 2899 |
+
"epoch": 0.3372727272727273,
|
| 2900 |
+
"grad_norm": 1.4489552247320478e+19,
|
| 2901 |
+
"learning_rate": 7.447465711034404e-07,
|
| 2902 |
+
"loss": 2.867232894897461,
|
| 2903 |
+
"step": 3710
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"epoch": 0.3381818181818182,
|
| 2907 |
+
"grad_norm": Infinity,
|
| 2908 |
+
"learning_rate": 7.43500351341541e-07,
|
| 2909 |
+
"loss": 2.8905174255371096,
|
| 2910 |
+
"step": 3720
|
| 2911 |
+
},
|
| 2912 |
+
{
|
| 2913 |
+
"epoch": 0.3390909090909091,
|
| 2914 |
+
"grad_norm": Infinity,
|
| 2915 |
+
"learning_rate": 7.422521454221989e-07,
|
| 2916 |
+
"loss": 2.171013069152832,
|
| 2917 |
+
"step": 3730
|
| 2918 |
+
},
|
| 2919 |
+
{
|
| 2920 |
+
"epoch": 0.34,
|
| 2921 |
+
"grad_norm": Infinity,
|
| 2922 |
+
"learning_rate": 7.410019635266459e-07,
|
| 2923 |
+
"loss": 2.721842384338379,
|
| 2924 |
+
"step": 3740
|
| 2925 |
+
},
|
| 2926 |
+
{
|
| 2927 |
+
"epoch": 0.3409090909090909,
|
| 2928 |
+
"grad_norm": 3.9303648421829673e+18,
|
| 2929 |
+
"learning_rate": 7.397498158522306e-07,
|
| 2930 |
+
"loss": 2.7071441650390624,
|
| 2931 |
+
"step": 3750
|
| 2932 |
+
},
|
| 2933 |
+
{
|
| 2934 |
+
"epoch": 0.3418181818181818,
|
| 2935 |
+
"grad_norm": Infinity,
|
| 2936 |
+
"learning_rate": 7.384957126123365e-07,
|
| 2937 |
+
"loss": 3.0133041381835937,
|
| 2938 |
+
"step": 3760
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 0.3427272727272727,
|
| 2942 |
+
"grad_norm": 7.447019192961729e+18,
|
| 2943 |
+
"learning_rate": 7.37239664036298e-07,
|
| 2944 |
+
"loss": 2.8034151077270506,
|
| 2945 |
+
"step": 3770
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 0.34363636363636363,
|
| 2949 |
+
"grad_norm": Infinity,
|
| 2950 |
+
"learning_rate": 7.359816803693166e-07,
|
| 2951 |
+
"loss": 2.6570461273193358,
|
| 2952 |
+
"step": 3780
|
| 2953 |
+
},
|
| 2954 |
+
{
|
| 2955 |
+
"epoch": 0.34454545454545454,
|
| 2956 |
+
"grad_norm": Infinity,
|
| 2957 |
+
"learning_rate": 7.347217718723783e-07,
|
| 2958 |
+
"loss": 3.0319595336914062,
|
| 2959 |
+
"step": 3790
|
| 2960 |
+
},
|
| 2961 |
+
{
|
| 2962 |
+
"epoch": 0.34545454545454546,
|
| 2963 |
+
"grad_norm": Infinity,
|
| 2964 |
+
"learning_rate": 7.334599488221689e-07,
|
| 2965 |
+
"loss": 2.714298439025879,
|
| 2966 |
+
"step": 3800
|
| 2967 |
+
},
|
| 2968 |
+
{
|
| 2969 |
+
"epoch": 0.34545454545454546,
|
| 2970 |
+
"eval_loss": 2.7910683155059814,
|
| 2971 |
+
"eval_runtime": 25.2334,
|
| 2972 |
+
"eval_samples_per_second": 5.826,
|
| 2973 |
+
"eval_steps_per_second": 1.466,
|
| 2974 |
+
"step": 3800
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 0.3463636363636364,
|
| 2978 |
+
"grad_norm": Infinity,
|
| 2979 |
+
"learning_rate": 7.321962215109906e-07,
|
| 2980 |
+
"loss": 3.015147590637207,
|
| 2981 |
+
"step": 3810
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 0.3472727272727273,
|
| 2985 |
+
"grad_norm": Infinity,
|
| 2986 |
+
"learning_rate": 7.309306002466787e-07,
|
| 2987 |
+
"loss": 3.0093734741210936,
|
| 2988 |
+
"step": 3820
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 0.3481818181818182,
|
| 2992 |
+
"grad_norm": Infinity,
|
| 2993 |
+
"learning_rate": 7.296630953525158e-07,
|
| 2994 |
+
"loss": 2.9403078079223635,
|
| 2995 |
+
"step": 3830
|
| 2996 |
+
},
|
| 2997 |
+
{
|
| 2998 |
+
"epoch": 0.3490909090909091,
|
| 2999 |
+
"grad_norm": Infinity,
|
| 3000 |
+
"learning_rate": 7.283937171671497e-07,
|
| 3001 |
+
"loss": 2.6926576614379885,
|
| 3002 |
+
"step": 3840
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 0.35,
|
| 3006 |
+
"grad_norm": Infinity,
|
| 3007 |
+
"learning_rate": 7.271224760445079e-07,
|
| 3008 |
+
"loss": 2.754840850830078,
|
| 3009 |
+
"step": 3850
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 0.3509090909090909,
|
| 3013 |
+
"grad_norm": Infinity,
|
| 3014 |
+
"learning_rate": 7.258493823537124e-07,
|
| 3015 |
+
"loss": 2.694375419616699,
|
| 3016 |
+
"step": 3860
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 0.3518181818181818,
|
| 3020 |
+
"grad_norm": Infinity,
|
| 3021 |
+
"learning_rate": 7.245744464789973e-07,
|
| 3022 |
+
"loss": 2.945112419128418,
|
| 3023 |
+
"step": 3870
|
| 3024 |
+
},
|
| 3025 |
+
{
|
| 3026 |
+
"epoch": 0.3527272727272727,
|
| 3027 |
+
"grad_norm": Infinity,
|
| 3028 |
+
"learning_rate": 7.232976788196221e-07,
|
| 3029 |
+
"loss": 2.8494308471679686,
|
| 3030 |
+
"step": 3880
|
| 3031 |
+
},
|
| 3032 |
+
{
|
| 3033 |
+
"epoch": 0.35363636363636364,
|
| 3034 |
+
"grad_norm": 1.6991103928231789e+19,
|
| 3035 |
+
"learning_rate": 7.220190897897877e-07,
|
| 3036 |
+
"loss": 2.9483663558959963,
|
| 3037 |
+
"step": 3890
|
| 3038 |
+
},
|
| 3039 |
+
{
|
| 3040 |
+
"epoch": 0.35454545454545455,
|
| 3041 |
+
"grad_norm": Infinity,
|
| 3042 |
+
"learning_rate": 7.207386898185515e-07,
|
| 3043 |
+
"loss": 2.6762250900268554,
|
| 3044 |
+
"step": 3900
|
| 3045 |
+
},
|
| 3046 |
+
{
|
| 3047 |
+
"epoch": 0.35454545454545455,
|
| 3048 |
+
"eval_loss": 2.7912540435791016,
|
| 3049 |
+
"eval_runtime": 25.2005,
|
| 3050 |
+
"eval_samples_per_second": 5.833,
|
| 3051 |
+
"eval_steps_per_second": 1.468,
|
| 3052 |
+
"step": 3900
|
| 3053 |
+
},
|
| 3054 |
+
{
|
| 3055 |
+
"epoch": 0.35545454545454547,
|
| 3056 |
+
"grad_norm": Infinity,
|
| 3057 |
+
"learning_rate": 7.194564893497419e-07,
|
| 3058 |
+
"loss": 2.797460174560547,
|
| 3059 |
+
"step": 3910
|
| 3060 |
+
},
|
| 3061 |
+
{
|
| 3062 |
+
"epoch": 0.3563636363636364,
|
| 3063 |
+
"grad_norm": Infinity,
|
| 3064 |
+
"learning_rate": 7.181724988418737e-07,
|
| 3065 |
+
"loss": 3.264780044555664,
|
| 3066 |
+
"step": 3920
|
| 3067 |
+
},
|
| 3068 |
+
{
|
| 3069 |
+
"epoch": 0.3572727272727273,
|
| 3070 |
+
"grad_norm": Infinity,
|
| 3071 |
+
"learning_rate": 7.168867287680627e-07,
|
| 3072 |
+
"loss": 2.6318002700805665,
|
| 3073 |
+
"step": 3930
|
| 3074 |
+
},
|
| 3075 |
+
{
|
| 3076 |
+
"epoch": 0.35818181818181816,
|
| 3077 |
+
"grad_norm": 3.8170838084406477e+18,
|
| 3078 |
+
"learning_rate": 7.155991896159392e-07,
|
| 3079 |
+
"loss": 2.9560144424438475,
|
| 3080 |
+
"step": 3940
|
| 3081 |
+
},
|
| 3082 |
+
{
|
| 3083 |
+
"epoch": 0.35909090909090907,
|
| 3084 |
+
"grad_norm": Infinity,
|
| 3085 |
+
"learning_rate": 7.143098918875642e-07,
|
| 3086 |
+
"loss": 2.273042678833008,
|
| 3087 |
+
"step": 3950
|
| 3088 |
+
},
|
| 3089 |
+
{
|
| 3090 |
+
"epoch": 0.36,
|
| 3091 |
+
"grad_norm": Infinity,
|
| 3092 |
+
"learning_rate": 7.130188460993426e-07,
|
| 3093 |
+
"loss": 2.5968145370483398,
|
| 3094 |
+
"step": 3960
|
| 3095 |
+
},
|
| 3096 |
+
{
|
| 3097 |
+
"epoch": 0.3609090909090909,
|
| 3098 |
+
"grad_norm": 1.2694314741464564e+19,
|
| 3099 |
+
"learning_rate": 7.117260627819376e-07,
|
| 3100 |
+
"loss": 2.8214059829711915,
|
| 3101 |
+
"step": 3970
|
| 3102 |
+
},
|
| 3103 |
+
{
|
| 3104 |
+
"epoch": 0.3618181818181818,
|
| 3105 |
+
"grad_norm": Infinity,
|
| 3106 |
+
"learning_rate": 7.104315524801848e-07,
|
| 3107 |
+
"loss": 2.538612174987793,
|
| 3108 |
+
"step": 3980
|
| 3109 |
+
},
|
| 3110 |
+
{
|
| 3111 |
+
"epoch": 0.36272727272727273,
|
| 3112 |
+
"grad_norm": Infinity,
|
| 3113 |
+
"learning_rate": 7.091353257530068e-07,
|
| 3114 |
+
"loss": 2.7980751037597655,
|
| 3115 |
+
"step": 3990
|
| 3116 |
+
},
|
| 3117 |
+
{
|
| 3118 |
+
"epoch": 0.36363636363636365,
|
| 3119 |
+
"grad_norm": Infinity,
|
| 3120 |
+
"learning_rate": 7.078373931733258e-07,
|
| 3121 |
+
"loss": 2.8797430038452148,
|
| 3122 |
+
"step": 4000
|
| 3123 |
+
},
|
| 3124 |
+
{
|
| 3125 |
+
"epoch": 0.36363636363636365,
|
| 3126 |
+
"eval_loss": 2.791391372680664,
|
| 3127 |
+
"eval_runtime": 25.1579,
|
| 3128 |
+
"eval_samples_per_second": 5.843,
|
| 3129 |
+
"eval_steps_per_second": 1.471,
|
| 3130 |
+
"step": 4000
|
| 3131 |
+
}
|
| 3132 |
+
],
|
| 3133 |
+
"logging_steps": 10,
|
| 3134 |
+
"max_steps": 11000,
|
| 3135 |
+
"num_input_tokens_seen": 0,
|
| 3136 |
+
"num_train_epochs": 9223372036854775807,
|
| 3137 |
+
"save_steps": 4000,
|
| 3138 |
+
"stateful_callbacks": {
|
| 3139 |
+
"TrainerControl": {
|
| 3140 |
+
"args": {
|
| 3141 |
+
"should_epoch_stop": false,
|
| 3142 |
+
"should_evaluate": false,
|
| 3143 |
+
"should_log": false,
|
| 3144 |
+
"should_save": true,
|
| 3145 |
+
"should_training_stop": false
|
| 3146 |
+
},
|
| 3147 |
+
"attributes": {}
|
| 3148 |
+
}
|
| 3149 |
+
},
|
| 3150 |
+
"total_flos": 4529923817472000.0,
|
| 3151 |
+
"train_batch_size": 1,
|
| 3152 |
+
"trial_name": null,
|
| 3153 |
+
"trial_params": null
|
| 3154 |
+
}
|
last-checkpoint/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:54fe93f6999fd45439c74e373e9d3db8f96eb3219246cb5cdf24b12c5b9c915e
|
| 3 |
+
size 5265
|