archive e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300 (2026-10-21)
Browse files- .gitattributes +14 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/hf/config.json +52 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/hf/export_manifest.json +16 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/hf/model.safetensors +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/lr_scheduler/.metadata +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/lr_scheduler/__0_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/lr_scheduler/__1_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/lr_scheduler/__2_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/lr_scheduler/__3_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/lr_scheduler/__4_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/lr_scheduler/__5_0.distcp +0 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/meta.json +81 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/.metadata +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__0_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__1_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__2_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__3_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__4_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__5_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/.metadata +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__0_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__1_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__2_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__3_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__4_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__5_0.distcp +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00000.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00001.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00002.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00003.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00004.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00005.pt +3 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/state_fingerprint_rank_00000.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/state_fingerprint_rank_00001.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/state_fingerprint_rank_00002.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/state_fingerprint_rank_00003.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/state_fingerprint_rank_00004.json +37 -0
- e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/state_fingerprint_rank_00005.json +37 -0
.gitattributes
CHANGED
|
@@ -858,3 +858,17 @@ e5_8b_800k_20260831/arms/e5-dense-lr2e3/checkpoints/iter_0026650/optimizer/__2_0
|
|
| 858 |
e5_8b_800k_20260831/arms/e5-dense-lr2e3/checkpoints/iter_0026650/optimizer/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 859 |
e5_8b_800k_20260831/arms/e5-dense-lr2e3/checkpoints/iter_0026650/optimizer/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 860 |
e5_8b_800k_20260831/arms/e5-dense-lr2e3/checkpoints/iter_0026650/optimizer/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 858 |
e5_8b_800k_20260831/arms/e5-dense-lr2e3/checkpoints/iter_0026650/optimizer/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 859 |
e5_8b_800k_20260831/arms/e5-dense-lr2e3/checkpoints/iter_0026650/optimizer/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 860 |
e5_8b_800k_20260831/arms/e5-dense-lr2e3/checkpoints/iter_0026650/optimizer/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 861 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/.metadata filter=lfs diff=lfs merge=lfs -text
|
| 862 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 863 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 864 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 865 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 866 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 867 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 868 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/.metadata filter=lfs diff=lfs merge=lfs -text
|
| 869 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 870 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 871 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 872 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 873 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__4_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 874 |
+
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__5_0.distcp filter=lfs diff=lfs merge=lfs -text
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/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_0053300/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": 53300
|
| 16 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/hf/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5dccebc9d6a4a093caa2e3bb0dff3283df38273b0be38d5e25501f741875fea2
|
| 3 |
+
size 8137392760
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/lr_scheduler/.metadata
ADDED
|
Binary file (3.45 kB). View file
|
|
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/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_0053300/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_0053300/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_0053300/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_0053300/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_0053300/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_0053300/meta.json
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eval/acc_0": 0.8199459736563047,
|
| 3 |
+
"eval/acc_1": 0.6863351443880302,
|
| 4 |
+
"eval/acc_10": 0.2038244037358133,
|
| 5 |
+
"eval/acc_11": 0.18480706185681334,
|
| 6 |
+
"eval/acc_12": 0.1699821702601132,
|
| 7 |
+
"eval/acc_13": 0.15541564890970153,
|
| 8 |
+
"eval/acc_14": 0.14355637220354592,
|
| 9 |
+
"eval/acc_2": 0.5796357951210147,
|
| 10 |
+
"eval/acc_3": 0.49415874974894486,
|
| 11 |
+
"eval/acc_4": 0.4244493213502723,
|
| 12 |
+
"eval/acc_5": 0.3670875501807307,
|
| 13 |
+
"eval/acc_6": 0.321728180483583,
|
| 14 |
+
"eval/acc_7": 0.28222304690689565,
|
| 15 |
+
"eval/acc_8": 0.2510223060375977,
|
| 16 |
+
"eval/acc_9": 0.2246629047149013,
|
| 17 |
+
"eval/accept_length": 4.614873078419408,
|
| 18 |
+
"eval/accepted_draft_tokens": 3.614873078419408,
|
| 19 |
+
"eval/avg_acc": 0.35435298208491733,
|
| 20 |
+
"eval/avg_loss": 2.364372269481664,
|
| 21 |
+
"eval/block_count": 251730,
|
| 22 |
+
"eval/domain/code/accept_length": 4.998924037013127,
|
| 23 |
+
"eval/domain/code/accepted_draft_tokens": 3.998924037013127,
|
| 24 |
+
"eval/domain/code/avg_acc": 0.38277010177147813,
|
| 25 |
+
"eval/domain/code/avg_loss": 2.040035375746443,
|
| 26 |
+
"eval/domain/code/block_count": 51117,
|
| 27 |
+
"eval/domain/code/prompt_count": 100,
|
| 28 |
+
"eval/domain/creative_writing/accept_length": 3.456102025152034,
|
| 29 |
+
"eval/domain/creative_writing/accepted_draft_tokens": 2.456102025152034,
|
| 30 |
+
"eval/domain/creative_writing/avg_acc": 0.2733787903970459,
|
| 31 |
+
"eval/domain/creative_writing/avg_loss": 3.0160360960251364,
|
| 32 |
+
"eval/domain/creative_writing/block_count": 50811,
|
| 33 |
+
"eval/domain/creative_writing/prompt_count": 100,
|
| 34 |
+
"eval/domain/factual_qa/accept_length": 3.682810230756818,
|
| 35 |
+
"eval/domain/factual_qa/accepted_draft_tokens": 2.682810230756818,
|
| 36 |
+
"eval/domain/factual_qa/avg_acc": 0.28934066637855443,
|
| 37 |
+
"eval/domain/factual_qa/avg_loss": 2.815722604195107,
|
| 38 |
+
"eval/domain/factual_qa/block_count": 49576,
|
| 39 |
+
"eval/domain/factual_qa/prompt_count": 100,
|
| 40 |
+
"eval/domain/general/accept_length": 4.345718539684679,
|
| 41 |
+
"eval/domain/general/accepted_draft_tokens": 3.3457185396846785,
|
| 42 |
+
"eval/domain/general/avg_acc": 0.336799160771924,
|
| 43 |
+
"eval/domain/general/avg_loss": 2.477760824127998,
|
| 44 |
+
"eval/domain/general/block_count": 49037,
|
| 45 |
+
"eval/domain/general/prompt_count": 100,
|
| 46 |
+
"eval/domain/math/accept_length": 6.542108656156596,
|
| 47 |
+
"eval/domain/math/accepted_draft_tokens": 5.542108656156596,
|
| 48 |
+
"eval/domain/math/avg_acc": 0.48543274098026085,
|
| 49 |
+
"eval/domain/math/avg_loss": 1.4990542729736265,
|
| 50 |
+
"eval/domain/math/block_count": 51189,
|
| 51 |
+
"eval/domain/math/prompt_count": 100,
|
| 52 |
+
"eval/macro5/accept_length": 4.6051326977526506,
|
| 53 |
+
"eval/macro5/accepted_draft_tokens": 3.6051326977526506,
|
| 54 |
+
"eval/macro5/avg_acc": 0.35354429205985266,
|
| 55 |
+
"eval/macro5/avg_loss": 2.3697218346136624,
|
| 56 |
+
"eval/ploss_0": 0.5678821638479864,
|
| 57 |
+
"eval/ploss_1": 1.157531910535018,
|
| 58 |
+
"eval/ploss_10": 3.9014429860444277,
|
| 59 |
+
"eval/ploss_11": 4.026896340517169,
|
| 60 |
+
"eval/ploss_12": 4.139056835383948,
|
| 61 |
+
"eval/ploss_13": 4.2385942508642716,
|
| 62 |
+
"eval/ploss_14": 4.329291628232699,
|
| 63 |
+
"eval/ploss_2": 1.6996520176299301,
|
| 64 |
+
"eval/ploss_3": 2.1695657859033908,
|
| 65 |
+
"eval/ploss_4": 2.570708983992997,
|
| 66 |
+
"eval/ploss_5": 2.8998796194722942,
|
| 67 |
+
"eval/ploss_6": 3.179907256969437,
|
| 68 |
+
"eval/ploss_7": 3.4142781762172776,
|
| 69 |
+
"eval/ploss_8": 3.601324995380662,
|
| 70 |
+
"eval/ploss_9": 3.7656611677283385,
|
| 71 |
+
"eval/prompt_count": 500,
|
| 72 |
+
"eval/wallclock_seconds": 20.348069375380874,
|
| 73 |
+
"global_step": 53300,
|
| 74 |
+
"inference_id": 53300,
|
| 75 |
+
"iteration": 53300,
|
| 76 |
+
"next_inference_id": 53300,
|
| 77 |
+
"next_step": 53300,
|
| 78 |
+
"step": 53300,
|
| 79 |
+
"timestamp": 1792567904.088045,
|
| 80 |
+
"world_size": 6
|
| 81 |
+
}
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/.metadata
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a2a3d51848ccb2bf3451989f0d10a02f58a32eb99e9649cfab6ed7e97714b97a
|
| 3 |
+
size 179737
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__0_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ead19a2c5f7ae338aab972a696ebaaaecc0a97106150e51174ae90793ddde9e5
|
| 3 |
+
size 1564178857
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__1_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28275062851899dd30a203880f446a7a2a5ab4efcce01dad0a1eccfe241194a1
|
| 3 |
+
size 1564178857
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__2_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac69d8a36b686b61d9142260e6ae0aa208a60117be489838081388146d9c6d13
|
| 3 |
+
size 1564178857
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__3_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8e59de732edef4e69aefe6ba91825d688c879365fa5441bff5beab7b671b64f
|
| 3 |
+
size 1564178857
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__4_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba08e8c70b2831d47df4152549eaebe2acc0be590db12243884d30179833d316
|
| 3 |
+
size 1564081167
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/model/__5_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a1dde0440060c68011a38fb7f7808e6a9b1865106b6fa1e4ba1dc29bc14fcec
|
| 3 |
+
size 1562441513
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/.metadata
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8eeedb562d8aa0f0c4020055e6f7a08d213db95fe324dd058798d9aeb788226
|
| 3 |
+
size 468667
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__0_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fe6c6f855b75128f3ff5108db5529f8d48adf7c5d1d1bf07520e3a2ea259cbb4
|
| 3 |
+
size 8139539753
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__1_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a6bbbaa313cbf0501f09404cfa59a39e63815bb62656bcd8e9ca09df99946fce
|
| 3 |
+
size 8139590655
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__2_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f368096187695cfc36cadecd327cf0afba46bfe0cd13a1cb50efcbc1e9e17e0
|
| 3 |
+
size 8139589266
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__3_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:24441ea5b61daf38e40cdf0e0763eb7759e67f058453301b562a740636d35f5d
|
| 3 |
+
size 8139592733
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__4_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9603ffde5782cf88b8934b9e35702edf45fe18751b51185510773f9376fbb9af
|
| 3 |
+
size 8139299663
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/optimizer/__5_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ceb595953ce0fdbf39764a72d21e11d6b282edc54bbfb7e151c2b89beb37a604
|
| 3 |
+
size 8129514781
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00000.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ddd83007599ef64a61684ea75f66598af93055ce989469d94ca00130d7ca6e1a
|
| 3 |
+
size 14709
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00001.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8332c752071c429e87d0eb739fce0bd228c9bbff969c1bc2f0353dc1424acf64
|
| 3 |
+
size 14773
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00002.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2abaab5f659f8f66346baf5ab0c46fa1456602f0d6038f8663ab4bd02aec57a4
|
| 3 |
+
size 14773
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00003.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1c25459b85a5312ee43b9da4ea489a4543395ff23580a653ad5978ff1bedb15
|
| 3 |
+
size 14709
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00004.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b3e907cfaa7fab57fe3bd17eece26916c291c774da7974bdf1656cfcd7201d2
|
| 3 |
+
size 14773
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/rng_rank_00005.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f05fd42a2c9a05d2b838e087f8795254b750a003907a4e51efb2fd08d71c35e
|
| 3 |
+
size 14709
|
e5_8b_800k_20260831/arms/e5-x4rand-sharedlr1e3-explr2e3-mos/checkpoints/iter_0053300/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": "c3173bbccf67fe588f8149e7f35b529438f1c6fa54192d3783307a103a949ffe",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "8e14f93edf4058419f0b981fd1dedf6e0bca01e23c412da6b833d8dc974d9a5d",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "533aafb3e3fe49f8fffeaa429f351e87fa2e7e85cf7f764039f9305a49f8ed95",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "174ad328fb80c709536ead031ca0e8e59b8f443bb3f0312424085146e83bc15a",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 53300,
|
| 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_0053300/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": "f0a6ab62cefb5c4f7fb1c49eb5c17ad50c806d50a37753f14c96cb8e278af181",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "8b0fde134eec61e3f1a77a4b6e4962142972e858510e5d0af1fda41e9f6b51c8",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "a557be92a55f1f5369f1dfa626090d1356fe472b97c3726ab04add0a34b831ee",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "174ad328fb80c709536ead031ca0e8e59b8f443bb3f0312424085146e83bc15a",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 53300,
|
| 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_0053300/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": "07dac5510594eac15b71fea50cd1e785a0c606f42a61b855fd59d74447dc46e1",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "587d2ce765db0e3b07ce6e109ea190f2aa7c053d827b9ceccb7748bdd3f7276e",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "ec9b6a313da9695edc673abbbf7302cacabf038d1b6978c1999e2fcfbf3607fb",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "174ad328fb80c709536ead031ca0e8e59b8f443bb3f0312424085146e83bc15a",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 53300,
|
| 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_0053300/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": "4530b4126f4129c7cff841db26d23e314bf34ae0c5422444ea17debcff7ab7ba",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "85a2d408b8aa7e2b24cd2d63e3ad8656bad57dc1c29661bee4e6079974f3009b",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "0c8db3bee422008dcca44cd3e3166070ec0555da6b5dd878847bd45ddbe9cf68",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "174ad328fb80c709536ead031ca0e8e59b8f443bb3f0312424085146e83bc15a",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 53300,
|
| 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_0053300/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": "c3cf346fca2c2d9cb99033bd3b4a06c1e41be562abc3afdbffe06ea73cc7ce4d",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "348f15eef7bc4398bc835f8c2b595e06dbc6239e1e1fadcdaa186c914928cafd",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "173d7eeb8b6ed346913f1532d28ff31b247f6f9a2a4b11768e9e36bde41e8144",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "174ad328fb80c709536ead031ca0e8e59b8f443bb3f0312424085146e83bc15a",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 53300,
|
| 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_0053300/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": "2c57e9ab713268fce495cf7d20f91c7eca41f5557cbe7481a26073ca081a988f",
|
| 11 |
+
"tensor_count": 144
|
| 12 |
+
},
|
| 13 |
+
"optimizer": {
|
| 14 |
+
"sampled_tensor_count": 288,
|
| 15 |
+
"sha256": "375893e314798e09dd57b25824e4e5925d8319d324270f06007470bb194bceb7",
|
| 16 |
+
"tensor_count": 512
|
| 17 |
+
},
|
| 18 |
+
"rng": {
|
| 19 |
+
"sampled_tensor_count": 0,
|
| 20 |
+
"sha256": "caf999251f2851d2bae74132973b678dfeadfa5b903ba979596078ac2d47aa81",
|
| 21 |
+
"tensor_count": 2
|
| 22 |
+
},
|
| 23 |
+
"scheduler": {
|
| 24 |
+
"sampled_tensor_count": 0,
|
| 25 |
+
"sha256": "174ad328fb80c709536ead031ca0e8e59b8f443bb3f0312424085146e83bc15a",
|
| 26 |
+
"tensor_count": 0
|
| 27 |
+
}
|
| 28 |
+
},
|
| 29 |
+
"global_step": 53300,
|
| 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 |
+
}
|