upload dd/25M/500Mtok/postprefixlm_pct100.json
Browse files
dd/25M/500Mtok/postprefixlm_pct100.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "dd",
|
| 3 |
+
"arch": {
|
| 4 |
+
"dim": 448,
|
| 5 |
+
"num_encoder_layers": 8,
|
| 6 |
+
"num_decoder_layers": 4
|
| 7 |
+
},
|
| 8 |
+
"arch_label": "25M",
|
| 9 |
+
"non_emb_params": 28913920,
|
| 10 |
+
"pretrain_ckpt": "pct100.pt",
|
| 11 |
+
"sft_objective": "prefixlm",
|
| 12 |
+
"sft_train_loss": 3.764062035698148,
|
| 13 |
+
"sft_train_time_s": 197.122398853302,
|
| 14 |
+
"sft_n_steps": 762,
|
| 15 |
+
"sft_mfu_pct": 1.3037123700343605,
|
| 16 |
+
"sft_tokens_target": 49997414,
|
| 17 |
+
"sft_tokens_actual": 49995776,
|
| 18 |
+
"sft_lr": 0.0002,
|
| 19 |
+
"sft_batch_size": 4,
|
| 20 |
+
"sft_grad_accum": 8,
|
| 21 |
+
"eval_loss_held_out": 3.875670748480922,
|
| 22 |
+
"eval_n_batches": 100,
|
| 23 |
+
"eval_n_valid_tokens": 749488,
|
| 24 |
+
"eval_elapsed_s": 5.171890020370483,
|
| 25 |
+
"wallclock_s": 209.00517463684082
|
| 26 |
+
}
|