Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
Job has been terminated due to a temporary spike in resource usage and may be restarted later.
Error code: JobManagerCrashedError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
kind string | created_at timestamp[s] | source_ckpt string | source_step int64 | source_seen_tok int64 | cfg dict | tie_weights bool | tokenizer_id string | tokenizer dict | vocab int64 | dblock_blocks int64 | steps int64 | batch_size int64 | block_size int64 | shared string | packages list | wall_sec float64 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
agillm4_dblock_bench_manifest_v1 | 2026-06-18T07:59:19 | /workspace/agillm4_4090_ckpts_active/pretrain_step02181113.pt | 2,181,113 | 19,278,130,176 | {
"d": 1280,
"layers": 28,
"heads": 20,
"rank": 160,
"tie_kv": true,
"moe_ffn": true,
"moe_experts": 2,
"moe_top_k": 1,
"moe_mlp_mult": 4,
"moe_shared_experts": 1,
"moe_shared_mlp_mult": 2
} | true | deepseek-ai/DeepSeek-V4-Pro | {
"payload_schema": 2,
"id": "deepseek-ai/DeepSeek-V4-Pro",
"has_tokenizer_json": true,
"has_tokenizer_bundle": true,
"special": {
"pad_token": "<｜end▁of▁sentence｜>",
"pad_token_id": 1,
"eos_token": "<｜end▁of▁sentence｜>",
"eos_token_id": 1,
"sep_token": null,
"sep_token_id": null,
... | 129,280 | 4 | 1 | 1 | 128 | /workspace/agillm41_side_rounds/side_cycle_20260618T075841Z/shared_frozen.pt | [
{
"worker_id": "geth",
"block_id": 1,
"layers": [
10,
11,
12,
13,
7,
8
],
"path": "/workspace/agillm41_side_rounds/side_cycle_20260618T075841Z/lease_geth_block1_agillm4bench.pt",
"bytes": 905654401
},
{
"worker_id": "mcp",
"block_id": 0,
"l... | 204.149 |
agillm4_dblock_bench_manifest_v1 | 2026-06-18T08:21:18 | /workspace/agillm4_4090_ckpts_active/pretrain_step02182564.pt | 2,182,564 | 19,327,162,368 | {
"d": 1280,
"layers": 28,
"heads": 20,
"rank": 160,
"tie_kv": true,
"moe_ffn": true,
"moe_experts": 2,
"moe_top_k": 1,
"moe_mlp_mult": 4,
"moe_shared_experts": 1,
"moe_shared_mlp_mult": 2
} | true | deepseek-ai/DeepSeek-V4-Pro | {
"payload_schema": 2,
"id": "deepseek-ai/DeepSeek-V4-Pro",
"has_tokenizer_json": true,
"has_tokenizer_bundle": true,
"special": {
"pad_token": "<｜end▁of▁sentence｜>",
"pad_token_id": 1,
"eos_token": "<｜end▁of▁sentence｜>",
"eos_token_id": 1,
"sep_token": null,
"sep_token_id": null,
... | 129,280 | 14 | 1 | 1 | 128 | /workspace/agillm41_side_rounds/side_cycle_20260618T082040Z/shared_frozen.pt | [
{
"worker_id": "geth",
"block_id": 6,
"layers": [
12,
13
],
"path": "/workspace/agillm41_side_rounds/side_cycle_20260618T082040Z/lease_geth_block6_agillm4bench.pt",
"bytes": 302158735
},
{
"worker_id": "mcp",
"block_id": 11,
"layers": [
22,
23
],
... | 267.63 |
No dataset card yet
- Downloads last month
- 1,381