upload dd/6.25M/1Btok/postprefixlm_pct100.json
Browse files
dd/6.25M/1Btok/postprefixlm_pct100.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "dd",
|
| 3 |
+
"arch": {
|
| 4 |
+
"dim": 192,
|
| 5 |
+
"num_encoder_layers": 10,
|
| 6 |
+
"num_decoder_layers": 5
|
| 7 |
+
},
|
| 8 |
+
"arch_label": "6.25M",
|
| 9 |
+
"non_emb_params": 6642048,
|
| 10 |
+
"pretrain_ckpt": "pct100.pt",
|
| 11 |
+
"sft_objective": "prefixlm",
|
| 12 |
+
"sft_train_loss": 4.035693138238288,
|
| 13 |
+
"sft_train_time_s": 514.6795392036438,
|
| 14 |
+
"sft_n_steps": 1525,
|
| 15 |
+
"sft_mfu_pct": 5.161648267627842,
|
| 16 |
+
"sft_tokens_target": 99994828,
|
| 17 |
+
"sft_tokens_actual": 99991552,
|
| 18 |
+
"sft_lr": 0.0002,
|
| 19 |
+
"sft_batch_size": 4,
|
| 20 |
+
"sft_grad_accum": 8,
|
| 21 |
+
"eval_loss_held_out": 4.112655288097716,
|
| 22 |
+
"eval_n_batches": 100,
|
| 23 |
+
"eval_n_valid_tokens": 749488,
|
| 24 |
+
"eval_elapsed_s": 12.071893692016602,
|
| 25 |
+
"wallclock_s": 531.0104389190674
|
| 26 |
+
}
|