archive e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325 (2026-10-20)
Browse files- .gitattributes +14 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/hf/config.json +52 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/hf/export_manifest.json +16 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/hf/model.safetensors +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/.metadata +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__0_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__1_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__2_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__3_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__4_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__5_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/meta.json +81 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/.metadata +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__0_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__1_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__2_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__3_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__4_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__5_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/.metadata +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__0_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__1_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__2_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__3_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__4_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__5_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00000.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00001.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00002.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00003.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00004.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00005.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00000.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00001.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00002.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00003.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00004.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00005.json +37 -0
.gitattributes
CHANGED
|
@@ -804,3 +804,17 @@ e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_002
|
|
| 804 |
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0026650/optimizer/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 805 |
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0026650/optimizer/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 806 |
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0026650/optimizer/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 804 |
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0026650/optimizer/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 805 |
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0026650/optimizer/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 806 |
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0026650/optimizer/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 807 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/.metadata filter=lfs diff=lfs merge=lfs -text
|
| 808 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 809 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 810 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 811 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 812 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 813 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 814 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/.metadata filter=lfs diff=lfs merge=lfs -text
|
| 815 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 816 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 817 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 818 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 819 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 820 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/hf/config.json
ADDED
|
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DFlashDraftModel"
|
| 4 |
+
],
|
| 5 |
+
"attention_backend": "fa4",
|
| 6 |
+
"domain_anchored_token_moe": false,
|
| 7 |
+
"domain_order": null,
|
| 8 |
+
"draft_vocab_size": 151936,
|
| 9 |
+
"head_dim": 128,
|
| 10 |
+
"hidden_size": 4096,
|
| 11 |
+
"intermediate_size": 12288,
|
| 12 |
+
"learned_router": false,
|
| 13 |
+
"mask_token_id": 151669,
|
| 14 |
+
"max_position_embeddings": 40960,
|
| 15 |
+
"model_type": "dflash",
|
| 16 |
+
"num_attention_heads": 32,
|
| 17 |
+
"num_domains": 1,
|
| 18 |
+
"num_experts": 4,
|
| 19 |
+
"num_hidden_layers": 5,
|
| 20 |
+
"num_key_value_heads": 8,
|
| 21 |
+
"num_specialists": 4,
|
| 22 |
+
"num_target_layers": 5,
|
| 23 |
+
"per_domain_context_proj": false,
|
| 24 |
+
"per_domain_layer": false,
|
| 25 |
+
"rms_norm_eps": 1e-06,
|
| 26 |
+
"rope_theta": 1000000.0,
|
| 27 |
+
"router_init_std": 0.0015625,
|
| 28 |
+
"router_reads_shared": false,
|
| 29 |
+
"router_top_k": 1,
|
| 30 |
+
"specialist_dispatch": "grouped",
|
| 31 |
+
"specialist_full_width": true,
|
| 32 |
+
"specialist_gate_mode": "confidence",
|
| 33 |
+
"specialist_init_std": 0.02,
|
| 34 |
+
"specialist_rank": 512,
|
| 35 |
+
"specialist_top_k": 2,
|
| 36 |
+
"specialist_top_k_tau": 0.9,
|
| 37 |
+
"specialist_up_init_std": 0.001,
|
| 38 |
+
"target_hidden_size": 4096,
|
| 39 |
+
"target_layer_ids": [
|
| 40 |
+
1,
|
| 41 |
+
9,
|
| 42 |
+
17,
|
| 43 |
+
25,
|
| 44 |
+
33
|
| 45 |
+
],
|
| 46 |
+
"target_num_hidden_layers": 36,
|
| 47 |
+
"tie_word_embeddings": false,
|
| 48 |
+
"token_moe": true,
|
| 49 |
+
"token_moe_shared": true,
|
| 50 |
+
"transformers_version": "4.57.1",
|
| 51 |
+
"vocab_size": 151936
|
| 52 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/hf/export_manifest.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"distributed_checkpoint_is_resume_authority": true,
|
| 3 |
+
"domain_anchored_token_moe": false,
|
| 4 |
+
"format": "dflash-hf-sglang",
|
| 5 |
+
"num_experts": 4,
|
| 6 |
+
"num_hidden_layers": 5,
|
| 7 |
+
"request_router_checkpoint_sha256": null,
|
| 8 |
+
"request_router_manifest_sha256": null,
|
| 9 |
+
"requires_external_request_router": false,
|
| 10 |
+
"router_semantics": "per-token-per-layer-hard-top1",
|
| 11 |
+
"router_top_k": 1,
|
| 12 |
+
"tensor_count": 128,
|
| 13 |
+
"tensor_schema_sha256": "sha256:f6a560a19d633906c790f87a9a4d76374266577061b2eba8052bddfd8b1178f6",
|
| 14 |
+
"token_moe": true,
|
| 15 |
+
"trainer_global_step": 13325
|
| 16 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/hf/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:04587adcc3e6f16f3441b30b3ad5fa3f9f94fe3904230ec6bdea676b71c488aa
|
| 3 |
+
size 8137392760
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/.metadata
ADDED
|
Binary file (3.45 kB). View file
|
|
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__0_0.distcp
ADDED
|
Binary file (3.78 kB). View file
|
|
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__1_0.distcp
ADDED
|
Binary file (3.78 kB). View file
|
|
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__2_0.distcp
ADDED
|
Binary file (2.52 kB). View file
|
|
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__3_0.distcp
ADDED
|
Binary file (2.52 kB). View file
|
|
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__4_0.distcp
ADDED
|
Binary file (2.52 kB). View file
|
|
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/lr_scheduler/__5_0.distcp
ADDED
|
Binary file (2.52 kB). View file
|
|
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/meta.json
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eval/acc_0": 0.7949151862639852,
|
| 3 |
+
"eval/acc_1": 0.6521759877739868,
|
| 4 |
+
"eval/acc_10": 0.1826655807772159,
|
| 5 |
+
"eval/acc_11": 0.16535137368903352,
|
| 6 |
+
"eval/acc_12": 0.1509966503053944,
|
| 7 |
+
"eval/acc_13": 0.13869211696549313,
|
| 8 |
+
"eval/acc_14": 0.12908067130997622,
|
| 9 |
+
"eval/acc_2": 0.5432271102202806,
|
| 10 |
+
"eval/acc_3": 0.45826768034093124,
|
| 11 |
+
"eval/acc_4": 0.3904350918595059,
|
| 12 |
+
"eval/acc_5": 0.3353461339208559,
|
| 13 |
+
"eval/acc_6": 0.29068280026333093,
|
| 14 |
+
"eval/acc_7": 0.2537331890805406,
|
| 15 |
+
"eval/acc_8": 0.22519039869108964,
|
| 16 |
+
"eval/acc_9": 0.20015008713841842,
|
| 17 |
+
"eval/accept_length": 4.212612720341449,
|
| 18 |
+
"eval/accepted_draft_tokens": 3.2126127203414483,
|
| 19 |
+
"eval/avg_acc": 0.32781102559540704,
|
| 20 |
+
"eval/avg_loss": 2.5248474573457167,
|
| 21 |
+
"eval/block_count": 251730,
|
| 22 |
+
"eval/domain/code/accept_length": 4.555079523446212,
|
| 23 |
+
"eval/domain/code/accepted_draft_tokens": 3.555079523446212,
|
| 24 |
+
"eval/domain/code/avg_acc": 0.3534314425906959,
|
| 25 |
+
"eval/domain/code/avg_loss": 2.2270649566051435,
|
| 26 |
+
"eval/domain/code/block_count": 51117,
|
| 27 |
+
"eval/domain/code/prompt_count": 100,
|
| 28 |
+
"eval/domain/creative_writing/accept_length": 3.129420794709807,
|
| 29 |
+
"eval/domain/creative_writing/accepted_draft_tokens": 2.129420794709807,
|
| 30 |
+
"eval/domain/creative_writing/avg_acc": 0.24982620911154174,
|
| 31 |
+
"eval/domain/creative_writing/avg_loss": 3.209087683430475,
|
| 32 |
+
"eval/domain/creative_writing/block_count": 50811,
|
| 33 |
+
"eval/domain/creative_writing/prompt_count": 100,
|
| 34 |
+
"eval/domain/factual_qa/accept_length": 3.376109407777957,
|
| 35 |
+
"eval/domain/factual_qa/accepted_draft_tokens": 2.376109407777957,
|
| 36 |
+
"eval/domain/factual_qa/avg_acc": 0.2682461091857324,
|
| 37 |
+
"eval/domain/factual_qa/avg_loss": 2.9475302150215237,
|
| 38 |
+
"eval/domain/factual_qa/block_count": 49576,
|
| 39 |
+
"eval/domain/factual_qa/prompt_count": 100,
|
| 40 |
+
"eval/domain/general/accept_length": 3.9798519503956755,
|
| 41 |
+
"eval/domain/general/accepted_draft_tokens": 2.9798519503956755,
|
| 42 |
+
"eval/domain/general/avg_acc": 0.3119899404657626,
|
| 43 |
+
"eval/domain/general/avg_loss": 2.6151256312997675,
|
| 44 |
+
"eval/domain/general/block_count": 49037,
|
| 45 |
+
"eval/domain/general/prompt_count": 100,
|
| 46 |
+
"eval/domain/math/accept_length": 5.978940788059934,
|
| 47 |
+
"eval/domain/math/accepted_draft_tokens": 4.978940788059934,
|
| 48 |
+
"eval/domain/math/avg_acc": 0.45182431445013876,
|
| 49 |
+
"eval/domain/math/avg_loss": 1.650534078437809,
|
| 50 |
+
"eval/domain/math/block_count": 51189,
|
| 51 |
+
"eval/domain/math/prompt_count": 100,
|
| 52 |
+
"eval/macro5/accept_length": 4.203880492877917,
|
| 53 |
+
"eval/macro5/accepted_draft_tokens": 3.203880492877917,
|
| 54 |
+
"eval/macro5/avg_acc": 0.32706360316077426,
|
| 55 |
+
"eval/macro5/avg_loss": 2.5298685129589438,
|
| 56 |
+
"eval/ploss_0": 0.6658939723344969,
|
| 57 |
+
"eval/ploss_1": 1.3148911808544033,
|
| 58 |
+
"eval/ploss_10": 4.059636767674526,
|
| 59 |
+
"eval/ploss_11": 4.181082920304082,
|
| 60 |
+
"eval/ploss_12": 4.2924833145483134,
|
| 61 |
+
"eval/ploss_13": 4.392126280727583,
|
| 62 |
+
"eval/ploss_14": 4.480050328472822,
|
| 63 |
+
"eval/ploss_2": 1.8859092006599874,
|
| 64 |
+
"eval/ploss_3": 2.36179575094549,
|
| 65 |
+
"eval/ploss_4": 2.758916680127039,
|
| 66 |
+
"eval/ploss_5": 3.0797410390101265,
|
| 67 |
+
"eval/ploss_6": 3.3515292546754276,
|
| 68 |
+
"eval/ploss_7": 3.5780989348969916,
|
| 69 |
+
"eval/ploss_8": 3.761979033871002,
|
| 70 |
+
"eval/ploss_9": 3.924510547664267,
|
| 71 |
+
"eval/prompt_count": 500,
|
| 72 |
+
"eval/wallclock_seconds": 20.32323621585965,
|
| 73 |
+
"global_step": 13325,
|
| 74 |
+
"inference_id": 13325,
|
| 75 |
+
"iteration": 13325,
|
| 76 |
+
"next_inference_id": 13325,
|
| 77 |
+
"next_step": 13325,
|
| 78 |
+
"step": 13325,
|
| 79 |
+
"timestamp": 1792401436.342078,
|
| 80 |
+
"world_size": 6
|
| 81 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/.metadata
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5eb87d67a17466554baeb59688690219d5b4b7f638e71edfc95681426734105f
|
| 3 |
+
size 179737
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__0_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c76933b44b78ade71ff26870f63af51b58eb2740abcd4a3dbc4986b71f60b5a
|
| 3 |
+
size 1564178857
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__1_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc822283e57a8b457ec4a0896ee03e40fb6c720a6e9da92ef637db21073a35d1
|
| 3 |
+
size 1564178857
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__2_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d4ba3e27ed3e597099319a65bd6c989864daf849b26c885a18e154166c87dd8
|
| 3 |
+
size 1564178857
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__3_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b37fcd8ce125cbc50a1572da0b94c08150c7496bcc18e7b131bc138a068e372
|
| 3 |
+
size 1564178857
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__4_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:166a58e2b0536ef4eacf18cc2a0d600eee7d1819d80883e957fad03343f8ce80
|
| 3 |
+
size 1564081167
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/model/__5_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1eb5299946962150f162983a9b80fb0806aa5c098dfaafdbf98a48fd9c5cea35
|
| 3 |
+
size 1562441513
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/.metadata
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7f5b31b31f595e79e84a477ca07b9b89ce6de0716f68bae47c1196ddc034efe3
|
| 3 |
+
size 468667
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__0_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:80f44bc33962c0a498f5b65bbcef78e7473f88e1b8eee226d9437c11d8df59f2
|
| 3 |
+
size 8139539753
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__1_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03d1c82cd2b4288a551fdc1dcb5d902d047eeaf691fcfdae1d64441bda131b7d
|
| 3 |
+
size 8139590655
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__2_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15abd8aed9fab24449d837c6396227a825fa5a195f089990ee457e843984bc58
|
| 3 |
+
size 8139589266
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__3_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78f8a0064689eb407fc9876e8e0d6a7ab1c9c19ff579691757b7e43a33f92008
|
| 3 |
+
size 8139592733
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__4_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08dc12d3c0b2a0e8ddbf7b94da7f6b4c78c664a6092fddd455943190c10e1e0e
|
| 3 |
+
size 8139299663
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/optimizer/__5_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6ca42068d4d20e0494209b5369b3b81e6f2c5b879333ea5ece0e3df8fe950ddf
|
| 3 |
+
size 8129514781
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00000.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d81ca07a9b568be446a2506e7ec07f46db08380b42f32c851290cdf28269fdbe
|
| 3 |
+
size 14709
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00001.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:303d37aa4f16765391d103f11f3d24d4b93983cd19997b3161983fbcc5330ffd
|
| 3 |
+
size 14773
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00002.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f454f22c5a8f64d35202f73ebe36d0638cb3ff4c63a98a13bbfbedb70bc243e4
|
| 3 |
+
size 14773
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00003.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:91357d92166832de6c0e837fb9b011f38455f1cd255919d59dd645b3c83d74b0
|
| 3 |
+
size 14709
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00004.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5bcb1f4684664a5e7dee87b313e3a3aa2100caa35b6f3903e7d6baa191d5c530
|
| 3 |
+
size 14773
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/rng_rank_00005.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a5148233d794a8cffe39c5333303576b07a2ebdc757314a646242384b84ae82e
|
| 3 |
+
size 14709
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00000.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"components": {
|
| 3 |
+
"data": {
|
| 4 |
+
"sampled_tensor_count": 0,
|
| 5 |
+
"sha256": "322594c96bcee57c77a5b7a138a9e672d5f3a23b976ba63295b5e030e84e4295",
|
| 6 |
+
"tensor_count": 0
|
| 7 |
+
},
|
| 8 |
+
"model": {
|
| 9 |
+
"sampled_tensor_count": 107,
|
| 10 |
+
"sha256": "8b19d8a0fbd3908de60235730de2394d41ead7cad54aa4910b1c2db909fe9c24",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "79c66e9523a81b269bb9f6bfabeb8594008a36b3b99ee69d57ac019c919f5ff7",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "bd4bc29abeb67fb4d1a64f5db98c216ebd8fab02ea4fc8d0e7b992d08d62a2ea",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "8007de2cc7db472bc62cc6faabed0e6eb58c6bd52486528cbbec96ba4530955b",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 13325,
|
| 30 |
+
"rank": 0,
|
| 31 |
+
"sampling": {
|
| 32 |
+
"online_batch_max_elements_per_tensor": null,
|
| 33 |
+
"state_max_elements_per_tensor": 8192
|
| 34 |
+
},
|
| 35 |
+
"schema_version": 1,
|
| 36 |
+
"world_size": 6
|
| 37 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00001.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"components": {
|
| 3 |
+
"data": {
|
| 4 |
+
"sampled_tensor_count": 0,
|
| 5 |
+
"sha256": "322594c96bcee57c77a5b7a138a9e672d5f3a23b976ba63295b5e030e84e4295",
|
| 6 |
+
"tensor_count": 0
|
| 7 |
+
},
|
| 8 |
+
"model": {
|
| 9 |
+
"sampled_tensor_count": 107,
|
| 10 |
+
"sha256": "c2bdeb61ead5916cff5b07eb85dc01694ba0d8128b910766aafdb03b22094585",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "d5f8278dc381a33e9b37fa6f7ad6fa3e064016ae39ce5f076beb9247f2d2d9db",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "8c93e26cb8de6859290b67aeccae8cbf1bec2c3c51e21e4781a2bf99c3a6b877",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "8007de2cc7db472bc62cc6faabed0e6eb58c6bd52486528cbbec96ba4530955b",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 13325,
|
| 30 |
+
"rank": 1,
|
| 31 |
+
"sampling": {
|
| 32 |
+
"online_batch_max_elements_per_tensor": null,
|
| 33 |
+
"state_max_elements_per_tensor": 8192
|
| 34 |
+
},
|
| 35 |
+
"schema_version": 1,
|
| 36 |
+
"world_size": 6
|
| 37 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00002.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"components": {
|
| 3 |
+
"data": {
|
| 4 |
+
"sampled_tensor_count": 0,
|
| 5 |
+
"sha256": "322594c96bcee57c77a5b7a138a9e672d5f3a23b976ba63295b5e030e84e4295",
|
| 6 |
+
"tensor_count": 0
|
| 7 |
+
},
|
| 8 |
+
"model": {
|
| 9 |
+
"sampled_tensor_count": 107,
|
| 10 |
+
"sha256": "fb23f9b04c68e25cb8c3f5afadccbdab0557be82ba6d3425420a686a1b62c141",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "e0e381e04994696cd22f38ec0304c93e9db0731b07119871307fd6adf70fac0d",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "e2e46043091edfe0f550a3f14d1d7068034847207e4421587a582544fce51618",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "8007de2cc7db472bc62cc6faabed0e6eb58c6bd52486528cbbec96ba4530955b",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 13325,
|
| 30 |
+
"rank": 2,
|
| 31 |
+
"sampling": {
|
| 32 |
+
"online_batch_max_elements_per_tensor": null,
|
| 33 |
+
"state_max_elements_per_tensor": 8192
|
| 34 |
+
},
|
| 35 |
+
"schema_version": 1,
|
| 36 |
+
"world_size": 6
|
| 37 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00003.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"components": {
|
| 3 |
+
"data": {
|
| 4 |
+
"sampled_tensor_count": 0,
|
| 5 |
+
"sha256": "322594c96bcee57c77a5b7a138a9e672d5f3a23b976ba63295b5e030e84e4295",
|
| 6 |
+
"tensor_count": 0
|
| 7 |
+
},
|
| 8 |
+
"model": {
|
| 9 |
+
"sampled_tensor_count": 107,
|
| 10 |
+
"sha256": "8387fe08ba6807c7380f5eab4ed93e8998b2e171c605cb90f56eb0e0a62f9fce",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "28be2903ed99b66dde80738124e843e9c87f77b1c97e14961090e0eb0cf7b121",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "c596caeb07c0b523cdd2c607f74094750c03e5ae6ef448e4a5f7226cba1a4ffe",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "8007de2cc7db472bc62cc6faabed0e6eb58c6bd52486528cbbec96ba4530955b",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 13325,
|
| 30 |
+
"rank": 3,
|
| 31 |
+
"sampling": {
|
| 32 |
+
"online_batch_max_elements_per_tensor": null,
|
| 33 |
+
"state_max_elements_per_tensor": 8192
|
| 34 |
+
},
|
| 35 |
+
"schema_version": 1,
|
| 36 |
+
"world_size": 6
|
| 37 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00004.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"components": {
|
| 3 |
+
"data": {
|
| 4 |
+
"sampled_tensor_count": 0,
|
| 5 |
+
"sha256": "322594c96bcee57c77a5b7a138a9e672d5f3a23b976ba63295b5e030e84e4295",
|
| 6 |
+
"tensor_count": 0
|
| 7 |
+
},
|
| 8 |
+
"model": {
|
| 9 |
+
"sampled_tensor_count": 107,
|
| 10 |
+
"sha256": "b1e3744535d2fcd9c396e3cba80f969ecdb72bbfa93f028d4f0bd2c082ed7ac8",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "ed2620ca6339184a830ed79acedad33a40e757ff9e8274606a1363d8316e0075",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "121bc3bc375477af3270e5cc1373ece296426db8ced237b969a613621f738d19",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "8007de2cc7db472bc62cc6faabed0e6eb58c6bd52486528cbbec96ba4530955b",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 13325,
|
| 30 |
+
"rank": 4,
|
| 31 |
+
"sampling": {
|
| 32 |
+
"online_batch_max_elements_per_tensor": null,
|
| 33 |
+
"state_max_elements_per_tensor": 8192
|
| 34 |
+
},
|
| 35 |
+
"schema_version": 1,
|
| 36 |
+
"world_size": 6
|
| 37 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0013325/state_fingerprint_rank_00005.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"components": {
|
| 3 |
+
"data": {
|
| 4 |
+
"sampled_tensor_count": 0,
|
| 5 |
+
"sha256": "322594c96bcee57c77a5b7a138a9e672d5f3a23b976ba63295b5e030e84e4295",
|
| 6 |
+
"tensor_count": 0
|
| 7 |
+
},
|
| 8 |
+
"model": {
|
| 9 |
+
"sampled_tensor_count": 107,
|
| 10 |
+
"sha256": "c17a0ef157ec07cdcca75ca16814fff6bafffca71a42d817ac25d57af0d0424f",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "e2f24ed1d733ce8879e92d9b69aac96597f1039b81bae30d8d7f1788349b7d0f",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "5a9ca45b27d87a5a4c57abed43bca004110d719115f8502f67ddda381004eef8",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "8007de2cc7db472bc62cc6faabed0e6eb58c6bd52486528cbbec96ba4530955b",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 13325,
|
| 30 |
+
"rank": 5,
|
| 31 |
+
"sampling": {
|
| 32 |
+
"online_batch_max_elements_per_tensor": null,
|
| 33 |
+
"state_max_elements_per_tensor": 8192
|
| 34 |
+
},
|
| 35 |
+
"schema_version": 1,
|
| 36 |
+
"world_size": 6
|
| 37 |
+
}
|