Text Generation
Transformers
TensorBoard
Safetensors
English
qwen3
byte-level
pretraining
symbolic
text-generation-inference
Instructions to use dotlabs/void.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use dotlabs/void.1 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="dotlabs/void.1")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("dotlabs/void.1") model = AutoModelForCausalLM.from_pretrained("dotlabs/void.1", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use dotlabs/void.1 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "dotlabs/void.1" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "dotlabs/void.1", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/dotlabs/void.1
- SGLang
How to use dotlabs/void.1 with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "dotlabs/void.1" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "dotlabs/void.1", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "dotlabs/void.1" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "dotlabs/void.1", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use dotlabs/void.1 with Docker Model Runner:
docker model run hf.co/dotlabs/void.1
Update TensorBoard training metrics
Browse files
benchmark_cache/performance.json
CHANGED
|
@@ -41,146 +41,146 @@
|
|
| 41 |
"attention_backend": "flash",
|
| 42 |
"vram_fraction": 0.98
|
| 43 |
},
|
| 44 |
-
"source": "
|
| 45 |
"requested_microbatch_ceiling": 90
|
| 46 |
},
|
| 47 |
-
"cache_reused":
|
| 48 |
"attention": "PyTorch FlashAttention (forced; forward/backward verified)",
|
| 49 |
"selected": {
|
| 50 |
"microbatch": 90,
|
| 51 |
"mode": "compiled",
|
| 52 |
-
"microbatch_seconds": 0.
|
| 53 |
-
"peak_gb": 89.
|
| 54 |
"fits_budget": true,
|
| 55 |
-
"estimated_update_seconds": 0.
|
| 56 |
-
"input_tokens_per_second":
|
| 57 |
},
|
| 58 |
"measurements": [
|
| 59 |
{
|
| 60 |
"microbatch": 8,
|
| 61 |
"mode": "eager",
|
| 62 |
-
"microbatch_seconds": 0.
|
| 63 |
"peak_gb": 14.764304637908936,
|
| 64 |
"fits_budget": true,
|
| 65 |
-
"estimated_update_seconds": 1.
|
| 66 |
-
"input_tokens_per_second":
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"microbatch": 16,
|
| 70 |
"mode": "eager",
|
| 71 |
-
"microbatch_seconds": 0.
|
| 72 |
"peak_gb": 28.084960460662842,
|
| 73 |
"fits_budget": true,
|
| 74 |
-
"estimated_update_seconds": 1.
|
| 75 |
-
"input_tokens_per_second":
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"microbatch": 32,
|
| 79 |
"mode": "eager",
|
| 80 |
-
"microbatch_seconds": 0.
|
| 81 |
"peak_gb": 54.827155113220215,
|
| 82 |
"fits_budget": true,
|
| 83 |
-
"estimated_update_seconds": 2.
|
| 84 |
-
"input_tokens_per_second":
|
| 85 |
},
|
| 86 |
{
|
| 87 |
"microbatch": 48,
|
| 88 |
"mode": "eager",
|
| 89 |
-
"microbatch_seconds": 1.
|
| 90 |
"peak_gb": 81.63183498382568,
|
| 91 |
"fits_budget": true,
|
| 92 |
-
"estimated_update_seconds": 2.
|
| 93 |
-
"input_tokens_per_second":
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"microbatch": 8,
|
| 97 |
"mode": "compiled",
|
| 98 |
-
"microbatch_seconds": 0.
|
| 99 |
"peak_gb": 9.113307476043701,
|
| 100 |
"fits_budget": true,
|
| 101 |
-
"estimated_update_seconds": 1.
|
| 102 |
-
"input_tokens_per_second":
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"microbatch": 16,
|
| 106 |
"mode": "compiled",
|
| 107 |
-
"microbatch_seconds": 0.
|
| 108 |
"peak_gb": 16.87287473678589,
|
| 109 |
"fits_budget": true,
|
| 110 |
-
"estimated_update_seconds": 0.
|
| 111 |
-
"input_tokens_per_second":
|
| 112 |
},
|
| 113 |
{
|
| 114 |
"microbatch": 32,
|
| 115 |
"mode": "compiled",
|
| 116 |
-
"microbatch_seconds": 0.
|
| 117 |
"peak_gb": 32.49641752243042,
|
| 118 |
"fits_budget": true,
|
| 119 |
-
"estimated_update_seconds": 0.
|
| 120 |
-
"input_tokens_per_second":
|
| 121 |
},
|
| 122 |
{
|
| 123 |
"microbatch": 48,
|
| 124 |
"mode": "compiled",
|
| 125 |
-
"microbatch_seconds": 0.
|
| 126 |
"peak_gb": 48.123157024383545,
|
| 127 |
"fits_budget": true,
|
| 128 |
-
"estimated_update_seconds": 0.
|
| 129 |
-
"input_tokens_per_second":
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"microbatch": 64,
|
| 133 |
"mode": "compiled",
|
| 134 |
-
"microbatch_seconds": 0.
|
| 135 |
"peak_gb": 63.757601737976074,
|
| 136 |
"fits_budget": true,
|
| 137 |
-
"estimated_update_seconds": 0.
|
| 138 |
-
"input_tokens_per_second":
|
| 139 |
},
|
| 140 |
{
|
| 141 |
"microbatch": 72,
|
| 142 |
"mode": "compiled",
|
| 143 |
-
"microbatch_seconds": 0.
|
| 144 |
"peak_gb": 71.65071392059326,
|
| 145 |
"fits_budget": true,
|
| 146 |
-
"estimated_update_seconds": 0.
|
| 147 |
-
"input_tokens_per_second":
|
| 148 |
},
|
| 149 |
{
|
| 150 |
"microbatch": 76,
|
| 151 |
"mode": "compiled",
|
| 152 |
-
"microbatch_seconds": 0.
|
| 153 |
"peak_gb": 75.5242338180542,
|
| 154 |
"fits_budget": true,
|
| 155 |
-
"estimated_update_seconds": 0.
|
| 156 |
-
"input_tokens_per_second":
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"microbatch": 80,
|
| 160 |
"mode": "compiled",
|
| 161 |
-
"microbatch_seconds": 0.
|
| 162 |
"peak_gb": 79.40062236785889,
|
| 163 |
"fits_budget": true,
|
| 164 |
-
"estimated_update_seconds": 0.
|
| 165 |
-
"input_tokens_per_second":
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"microbatch": 86,
|
| 169 |
"mode": "compiled",
|
| 170 |
-
"microbatch_seconds": 0.
|
| 171 |
"peak_gb": 85.4687147140503,
|
| 172 |
"fits_budget": true,
|
| 173 |
-
"estimated_update_seconds": 0.
|
| 174 |
-
"input_tokens_per_second":
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"microbatch": 90,
|
| 178 |
"mode": "compiled",
|
| 179 |
-
"microbatch_seconds": 0.
|
| 180 |
"peak_gb": 89.3772840499878,
|
| 181 |
"fits_budget": true,
|
| 182 |
-
"estimated_update_seconds": 0.
|
| 183 |
-
"input_tokens_per_second":
|
| 184 |
}
|
| 185 |
],
|
| 186 |
"failures": [
|
|
@@ -218,6 +218,5 @@
|
|
| 218 |
"selection_policy": "max_batch",
|
| 219 |
"budget_gb": 92.52170804977416,
|
| 220 |
"reserve_gb": 2.1791439056396484,
|
| 221 |
-
"note": "Forward/backward throughput; excludes optimizer, data, logging and checkpoint time."
|
| 222 |
-
"legacy_seed_verified": false
|
| 223 |
}
|
|
|
|
| 41 |
"attention_backend": "flash",
|
| 42 |
"vram_fraction": 0.98
|
| 43 |
},
|
| 44 |
+
"source": "d52752a64023cf29dc386d43835cf3173c5318ee093c3b375a812733c74023a1",
|
| 45 |
"requested_microbatch_ceiling": 90
|
| 46 |
},
|
| 47 |
+
"cache_reused": false,
|
| 48 |
"attention": "PyTorch FlashAttention (forced; forward/backward verified)",
|
| 49 |
"selected": {
|
| 50 |
"microbatch": 90,
|
| 51 |
"mode": "compiled",
|
| 52 |
+
"microbatch_seconds": 0.9299377820000245,
|
| 53 |
+
"peak_gb": 89.3772840499878,
|
| 54 |
"fits_budget": true,
|
| 55 |
+
"estimated_update_seconds": 0.9299377820000245,
|
| 56 |
+
"input_tokens_per_second": 198206.80863571487
|
| 57 |
},
|
| 58 |
"measurements": [
|
| 59 |
{
|
| 60 |
"microbatch": 8,
|
| 61 |
"mode": "eager",
|
| 62 |
+
"microbatch_seconds": 0.15785124100000303,
|
| 63 |
"peak_gb": 14.764304637908936,
|
| 64 |
"fits_budget": true,
|
| 65 |
+
"estimated_update_seconds": 1.7943123250000212,
|
| 66 |
+
"input_tokens_per_second": 103793.92582665654
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"microbatch": 16,
|
| 70 |
"mode": "eager",
|
| 71 |
+
"microbatch_seconds": 0.3454984619999948,
|
| 72 |
"peak_gb": 28.084960460662842,
|
| 73 |
"fits_budget": true,
|
| 74 |
+
"estimated_update_seconds": 1.92592258199997,
|
| 75 |
+
"input_tokens_per_second": 94842.67979172798
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"microbatch": 32,
|
| 79 |
"mode": "eager",
|
| 80 |
+
"microbatch_seconds": 0.7818738260000089,
|
| 81 |
"peak_gb": 54.827155113220215,
|
| 82 |
"fits_budget": true,
|
| 83 |
+
"estimated_update_seconds": 2.1863095480000254,
|
| 84 |
+
"input_tokens_per_second": 83819.15063620413
|
| 85 |
},
|
| 86 |
{
|
| 87 |
"microbatch": 48,
|
| 88 |
"mode": "eager",
|
| 89 |
+
"microbatch_seconds": 1.171907979999986,
|
| 90 |
"peak_gb": 81.63183498382568,
|
| 91 |
"fits_budget": true,
|
| 92 |
+
"estimated_update_seconds": 2.1970569614999818,
|
| 93 |
+
"input_tokens_per_second": 83883.71926608194
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"microbatch": 8,
|
| 97 |
"mode": "compiled",
|
| 98 |
+
"microbatch_seconds": 0.0880895730000475,
|
| 99 |
"peak_gb": 9.113307476043701,
|
| 100 |
"fits_budget": true,
|
| 101 |
+
"estimated_update_seconds": 1.0067526180005189,
|
| 102 |
+
"input_tokens_per_second": 185992.50106469658
|
| 103 |
},
|
| 104 |
{
|
| 105 |
"microbatch": 16,
|
| 106 |
"mode": "compiled",
|
| 107 |
+
"microbatch_seconds": 0.1701795650000122,
|
| 108 |
"peak_gb": 16.87287473678589,
|
| 109 |
"fits_budget": true,
|
| 110 |
+
"estimated_update_seconds": 0.9559394930000451,
|
| 111 |
+
"input_tokens_per_second": 192549.55787434086
|
| 112 |
},
|
| 113 |
{
|
| 114 |
"microbatch": 32,
|
| 115 |
"mode": "compiled",
|
| 116 |
+
"microbatch_seconds": 0.34178767000003063,
|
| 117 |
"peak_gb": 32.49641752243042,
|
| 118 |
"fits_budget": true,
|
| 119 |
+
"estimated_update_seconds": 0.950241911000063,
|
| 120 |
+
"input_tokens_per_second": 191744.77534544803
|
| 121 |
},
|
| 122 |
{
|
| 123 |
"microbatch": 48,
|
| 124 |
"mode": "compiled",
|
| 125 |
+
"microbatch_seconds": 0.5061955770000282,
|
| 126 |
"peak_gb": 48.123157024383545,
|
| 127 |
"fits_budget": true,
|
| 128 |
+
"estimated_update_seconds": 0.9457143725000208,
|
| 129 |
+
"input_tokens_per_second": 194201.61784620755
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"microbatch": 64,
|
| 133 |
"mode": "compiled",
|
| 134 |
+
"microbatch_seconds": 0.6595429280000076,
|
| 135 |
"peak_gb": 63.757601737976074,
|
| 136 |
"fits_budget": true,
|
| 137 |
+
"estimated_update_seconds": 0.9208904890000156,
|
| 138 |
+
"input_tokens_per_second": 198731.56762890573
|
| 139 |
},
|
| 140 |
{
|
| 141 |
"microbatch": 72,
|
| 142 |
"mode": "compiled",
|
| 143 |
+
"microbatch_seconds": 0.741346276999991,
|
| 144 |
"peak_gb": 71.65071392059326,
|
| 145 |
"fits_budget": true,
|
| 146 |
+
"estimated_update_seconds": 0.9243867199999727,
|
| 147 |
+
"input_tokens_per_second": 198903.00197730918
|
| 148 |
},
|
| 149 |
{
|
| 150 |
"microbatch": 76,
|
| 151 |
"mode": "compiled",
|
| 152 |
+
"microbatch_seconds": 0.7816785220000497,
|
| 153 |
"peak_gb": 75.5242338180542,
|
| 154 |
"fits_budget": true,
|
| 155 |
+
"estimated_update_seconds": 0.9251490005000278,
|
| 156 |
+
"input_tokens_per_second": 199120.22093398403
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"microbatch": 80,
|
| 160 |
"mode": "compiled",
|
| 161 |
+
"microbatch_seconds": 0.8258502979999776,
|
| 162 |
"peak_gb": 79.40062236785889,
|
| 163 |
"fits_budget": true,
|
| 164 |
+
"estimated_update_seconds": 0.9283437199999867,
|
| 165 |
+
"input_tokens_per_second": 198389.46646478592
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"microbatch": 86,
|
| 169 |
"mode": "compiled",
|
| 170 |
+
"microbatch_seconds": 0.8883623469999975,
|
| 171 |
"peak_gb": 85.4687147140503,
|
| 172 |
"fits_budget": true,
|
| 173 |
+
"estimated_update_seconds": 0.9419139469999607,
|
| 174 |
+
"input_tokens_per_second": 198261.44207347915
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"microbatch": 90,
|
| 178 |
"mode": "compiled",
|
| 179 |
+
"microbatch_seconds": 0.9299377820000245,
|
| 180 |
"peak_gb": 89.3772840499878,
|
| 181 |
"fits_budget": true,
|
| 182 |
+
"estimated_update_seconds": 0.9299377820000245,
|
| 183 |
+
"input_tokens_per_second": 198206.80863571487
|
| 184 |
}
|
| 185 |
],
|
| 186 |
"failures": [
|
|
|
|
| 218 |
"selection_policy": "max_batch",
|
| 219 |
"budget_gb": 92.52170804977416,
|
| 220 |
"reserve_gb": 2.1791439056396484,
|
| 221 |
+
"note": "Forward/backward throughput; excludes optimizer, data, logging and checkpoint time."
|
|
|
|
| 222 |
}
|
benchmark_cache/shard_scheduler.json
CHANGED
|
@@ -22,8 +22,8 @@
|
|
| 22 |
"tokens_per_second": 598315.063954628,
|
| 23 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 24 |
"estimated_pages_for_buffer": 1,
|
| 25 |
-
"required_token_positions_per_second":
|
| 26 |
-
"required_pages_per_second_upper_estimate": 0.
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"source": "rewrite6",
|
|
@@ -36,8 +36,8 @@
|
|
| 36 |
"tokens_per_second": 8533.957213384612,
|
| 37 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 38 |
"estimated_pages_for_buffer": 1,
|
| 39 |
-
"required_token_positions_per_second":
|
| 40 |
-
"required_pages_per_second_upper_estimate": 0.
|
| 41 |
},
|
| 42 |
{
|
| 43 |
"source": "openmath",
|
|
@@ -50,8 +50,8 @@
|
|
| 50 |
"tokens_per_second": 380144.35761751566,
|
| 51 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 52 |
"estimated_pages_for_buffer": 1,
|
| 53 |
-
"required_token_positions_per_second":
|
| 54 |
-
"required_pages_per_second_upper_estimate": 0.
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"source": "ultra_qa",
|
|
@@ -64,8 +64,8 @@
|
|
| 64 |
"tokens_per_second": 43742.16556775435,
|
| 65 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 66 |
"estimated_pages_for_buffer": 1,
|
| 67 |
-
"required_token_positions_per_second":
|
| 68 |
-
"required_pages_per_second_upper_estimate": 0.
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"source": "ultra_style",
|
|
@@ -78,8 +78,8 @@
|
|
| 78 |
"tokens_per_second": 22018.86276523721,
|
| 79 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 80 |
"estimated_pages_for_buffer": 1,
|
| 81 |
-
"required_token_positions_per_second":
|
| 82 |
-
"required_pages_per_second_upper_estimate": 0.
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"source": "cortex",
|
|
@@ -91,9 +91,9 @@
|
|
| 91 |
"read_seconds": 0.04061180899998362,
|
| 92 |
"tokens_per_second": 910315.184299418,
|
| 93 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 94 |
-
"estimated_pages_for_buffer":
|
| 95 |
-
"required_token_positions_per_second":
|
| 96 |
-
"required_pages_per_second_upper_estimate": 0.
|
| 97 |
}
|
| 98 |
],
|
| 99 |
"gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
|
|
@@ -155,187 +155,30 @@
|
|
| 155 |
"shard_benchmark_scope": "Local packing/read measurements; upload capacity measured only on real data commits. Targets larger than available pages are sample-limited.",
|
| 156 |
"estimated_fresh_generation_batch_seconds": 2.558590814491007,
|
| 157 |
"selected": {
|
| 158 |
-
"prefetch_batches":
|
| 159 |
"prefetch_capacity_limit": 512,
|
| 160 |
"batch_memory_estimate_bytes": 720750,
|
| 161 |
-
"producer_batch_seconds":
|
| 162 |
-
"producer_p95_seconds":
|
| 163 |
-
"gpu_update_seconds": 0.
|
| 164 |
-
"producer_batches_per_second":
|
| 165 |
-
"gpu_batches_per_second": 1.
|
| 166 |
-
"producer_headroom":
|
| 167 |
-
"sustainable":
|
| 168 |
-
"buffer_seconds": 60.
|
| 169 |
-
"buffer_drain_seconds":
|
| 170 |
"shard_target_mib": 16,
|
| 171 |
-
"upload_pages_per_commit":
|
| 172 |
-
"max_pending_pages":
|
| 173 |
-
"generation_bytes_per_second":
|
| 174 |
-
"upload_bytes_per_second":
|
| 175 |
-
"upload_sustainable":
|
| 176 |
"logical_page_rows": 4096,
|
| 177 |
"logical_cortex_page_rows": 128,
|
| 178 |
-
"estimated_full_shards_per_hour":
|
| 179 |
"selection_policy": "Smallest shard within 90% of measured local peak, enlarged to amortize 30 seconds of measured page production; bounded to 256 MiB."
|
| 180 |
},
|
| 181 |
-
"upload_measurements": [
|
| 182 |
-
{
|
| 183 |
-
"seconds": 1.296470003999957,
|
| 184 |
-
"bytes": 22520,
|
| 185 |
-
"pages": 1,
|
| 186 |
-
"shards": 1
|
| 187 |
-
},
|
| 188 |
-
{
|
| 189 |
-
"seconds": 1.5001352909998786,
|
| 190 |
-
"bytes": 11915224,
|
| 191 |
-
"pages": 36,
|
| 192 |
-
"shards": 3
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"seconds": 1.1385679780000828,
|
| 196 |
-
"bytes": 2605060,
|
| 197 |
-
"pages": 36,
|
| 198 |
-
"shards": 2
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"seconds": 1.2088590650000697,
|
| 202 |
-
"bytes": 9710127,
|
| 203 |
-
"pages": 36,
|
| 204 |
-
"shards": 3
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"seconds": 0.972143672999664,
|
| 208 |
-
"bytes": 6102938,
|
| 209 |
-
"pages": 35,
|
| 210 |
-
"shards": 2
|
| 211 |
-
},
|
| 212 |
-
{
|
| 213 |
-
"seconds": 0.7959529229997315,
|
| 214 |
-
"bytes": 2320780,
|
| 215 |
-
"pages": 28,
|
| 216 |
-
"shards": 2
|
| 217 |
-
},
|
| 218 |
-
{
|
| 219 |
-
"seconds": 1.0472984960001668,
|
| 220 |
-
"bytes": 9808086,
|
| 221 |
-
"pages": 37,
|
| 222 |
-
"shards": 3
|
| 223 |
-
},
|
| 224 |
-
{
|
| 225 |
-
"seconds": 1.0871454139996786,
|
| 226 |
-
"bytes": 13468906,
|
| 227 |
-
"pages": 38,
|
| 228 |
-
"shards": 4
|
| 229 |
-
},
|
| 230 |
-
{
|
| 231 |
-
"seconds": 0.8225870739997845,
|
| 232 |
-
"bytes": 1093468,
|
| 233 |
-
"pages": 34,
|
| 234 |
-
"shards": 1
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"seconds": 1.3778674129998763,
|
| 238 |
-
"bytes": 6041724,
|
| 239 |
-
"pages": 32,
|
| 240 |
-
"shards": 2
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"seconds": 1.0736160129999917,
|
| 244 |
-
"bytes": 6799126,
|
| 245 |
-
"pages": 32,
|
| 246 |
-
"shards": 3
|
| 247 |
-
},
|
| 248 |
-
{
|
| 249 |
-
"seconds": 1.1032466139999997,
|
| 250 |
-
"bytes": 6317020,
|
| 251 |
-
"pages": 35,
|
| 252 |
-
"shards": 2
|
| 253 |
-
},
|
| 254 |
-
{
|
| 255 |
-
"seconds": 1.0759855689998403,
|
| 256 |
-
"bytes": 6102397,
|
| 257 |
-
"pages": 35,
|
| 258 |
-
"shards": 2
|
| 259 |
-
},
|
| 260 |
-
{
|
| 261 |
-
"seconds": 1.0165042769999673,
|
| 262 |
-
"bytes": 8195972,
|
| 263 |
-
"pages": 32,
|
| 264 |
-
"shards": 3
|
| 265 |
-
},
|
| 266 |
-
{
|
| 267 |
-
"seconds": 0.9696753220000573,
|
| 268 |
-
"bytes": 10144942,
|
| 269 |
-
"pages": 38,
|
| 270 |
-
"shards": 3
|
| 271 |
-
},
|
| 272 |
-
{
|
| 273 |
-
"seconds": 1.035786852000001,
|
| 274 |
-
"bytes": 2660447,
|
| 275 |
-
"pages": 37,
|
| 276 |
-
"shards": 2
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"seconds": 1.2393170549999013,
|
| 280 |
-
"bytes": 5990595,
|
| 281 |
-
"pages": 27,
|
| 282 |
-
"shards": 2
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"seconds": 0.8981117179996545,
|
| 286 |
-
"bytes": 970766,
|
| 287 |
-
"pages": 28,
|
| 288 |
-
"shards": 1
|
| 289 |
-
},
|
| 290 |
-
{
|
| 291 |
-
"seconds": 1.1719487760001357,
|
| 292 |
-
"bytes": 12751643,
|
| 293 |
-
"pages": 48,
|
| 294 |
-
"shards": 5
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"seconds": 1.1927614570004152,
|
| 298 |
-
"bytes": 11797840,
|
| 299 |
-
"pages": 30,
|
| 300 |
-
"shards": 3
|
| 301 |
-
},
|
| 302 |
-
{
|
| 303 |
-
"seconds": 1.1542902260007395,
|
| 304 |
-
"bytes": 2688701,
|
| 305 |
-
"pages": 36,
|
| 306 |
-
"shards": 2
|
| 307 |
-
},
|
| 308 |
-
{
|
| 309 |
-
"seconds": 0.9894740659992749,
|
| 310 |
-
"bytes": 5828242,
|
| 311 |
-
"pages": 28,
|
| 312 |
-
"shards": 2
|
| 313 |
-
},
|
| 314 |
-
{
|
| 315 |
-
"seconds": 1.3037635500004399,
|
| 316 |
-
"bytes": 10399515,
|
| 317 |
-
"pages": 37,
|
| 318 |
-
"shards": 3
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"seconds": 1.1070935670004474,
|
| 322 |
-
"bytes": 2763911,
|
| 323 |
-
"pages": 37,
|
| 324 |
-
"shards": 2
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"seconds": 1.3125206449994948,
|
| 328 |
-
"bytes": 6101597,
|
| 329 |
-
"pages": 27,
|
| 330 |
-
"shards": 2
|
| 331 |
-
},
|
| 332 |
-
{
|
| 333 |
-
"seconds": 0.9716803040000741,
|
| 334 |
-
"bytes": 6849385,
|
| 335 |
-
"pages": 34,
|
| 336 |
-
"shards": 2
|
| 337 |
-
}
|
| 338 |
-
],
|
| 339 |
"cache_key": {
|
| 340 |
"version": 1,
|
| 341 |
"kind": "data",
|
|
@@ -378,7 +221,7 @@
|
|
| 378 |
},
|
| 379 |
"cache_reused": true,
|
| 380 |
"resume_probe_seconds": [
|
| 381 |
-
|
| 382 |
-
0.
|
| 383 |
]
|
| 384 |
}
|
|
|
|
| 22 |
"tokens_per_second": 598315.063954628,
|
| 23 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 24 |
"estimated_pages_for_buffer": 1,
|
| 25 |
+
"required_token_positions_per_second": 79282.72345428595,
|
| 26 |
+
"required_pages_per_second_upper_estimate": 0.005484962769830832
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"source": "rewrite6",
|
|
|
|
| 36 |
"tokens_per_second": 8533.957213384612,
|
| 37 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 38 |
"estimated_pages_for_buffer": 1,
|
| 39 |
+
"required_token_positions_per_second": 9910.340431785744,
|
| 40 |
+
"required_pages_per_second_upper_estimate": 0.004149252587347242
|
| 41 |
},
|
| 42 |
{
|
| 43 |
"source": "openmath",
|
|
|
|
| 50 |
"tokens_per_second": 380144.35761751566,
|
| 51 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 52 |
"estimated_pages_for_buffer": 1,
|
| 53 |
+
"required_token_positions_per_second": 19820.680863571488,
|
| 54 |
+
"required_pages_per_second_upper_estimate": 0.004128125117377078
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"source": "ultra_qa",
|
|
|
|
| 64 |
"tokens_per_second": 43742.16556775435,
|
| 65 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 66 |
"estimated_pages_for_buffer": 1,
|
| 67 |
+
"required_token_positions_per_second": 29731.02129535723,
|
| 68 |
+
"required_pages_per_second_upper_estimate": 0.0017823293300156986
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"source": "ultra_style",
|
|
|
|
| 78 |
"tokens_per_second": 22018.86276523721,
|
| 79 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 80 |
"estimated_pages_for_buffer": 1,
|
| 81 |
+
"required_token_positions_per_second": 29731.02129535723,
|
| 82 |
+
"required_pages_per_second_upper_estimate": 0.002311456132771857
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"source": "cortex",
|
|
|
|
| 91 |
"read_seconds": 0.04061180899998362,
|
| 92 |
"tokens_per_second": 910315.184299418,
|
| 93 |
"note": "Bounded fresh sample; full-page latency is an extrapolation.",
|
| 94 |
+
"estimated_pages_for_buffer": 24,
|
| 95 |
+
"required_token_positions_per_second": 29731.02129535723,
|
| 96 |
+
"required_pages_per_second_upper_estimate": 0.38135761849331373
|
| 97 |
}
|
| 98 |
],
|
| 99 |
"gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
|
|
|
|
| 155 |
"shard_benchmark_scope": "Local packing/read measurements; upload capacity measured only on real data commits. Targets larger than available pages are sample-limited.",
|
| 156 |
"estimated_fresh_generation_batch_seconds": 2.558590814491007,
|
| 157 |
"selected": {
|
| 158 |
+
"prefetch_batches": 65,
|
| 159 |
"prefetch_capacity_limit": 512,
|
| 160 |
"batch_memory_estimate_bytes": 720750,
|
| 161 |
+
"producer_batch_seconds": 1.304301914000007,
|
| 162 |
+
"producer_p95_seconds": 2.591754088000016,
|
| 163 |
+
"gpu_update_seconds": 0.9299377820000245,
|
| 164 |
+
"producer_batches_per_second": 0.7666936537210316,
|
| 165 |
+
"gpu_batches_per_second": 1.0753407586573074,
|
| 166 |
+
"producer_headroom": 0.7129773958148309,
|
| 167 |
+
"sustainable": false,
|
| 168 |
+
"buffer_seconds": 60.445955830001594,
|
| 169 |
+
"buffer_drain_seconds": 210.5964998875337,
|
| 170 |
"shard_target_mib": 16,
|
| 171 |
+
"upload_pages_per_commit": 7,
|
| 172 |
+
"max_pending_pages": 1769,
|
| 173 |
+
"generation_bytes_per_second": 0,
|
| 174 |
+
"upload_bytes_per_second": null,
|
| 175 |
+
"upload_sustainable": null,
|
| 176 |
"logical_page_rows": 4096,
|
| 177 |
"logical_cortex_page_rows": 128,
|
| 178 |
+
"estimated_full_shards_per_hour": 0.0,
|
| 179 |
"selection_policy": "Smallest shard within 90% of measured local peak, enlarged to amortize 30 seconds of measured page production; bounded to 256 MiB."
|
| 180 |
},
|
| 181 |
+
"upload_measurements": [],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 182 |
"cache_key": {
|
| 183 |
"version": 1,
|
| 184 |
"kind": "data",
|
|
|
|
| 221 |
},
|
| 222 |
"cache_reused": true,
|
| 223 |
"resume_probe_seconds": [
|
| 224 |
+
2.591754088000016,
|
| 225 |
+
0.016849739999997837
|
| 226 |
]
|
| 227 |
}
|
runs/void-byte-v1/events.out.tfevents.1791039222.f2a393d4-e1e4-407b-a6aa-034aa2ada5ac-vwfgg.353.0.1791039222312944374-63c10d34
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:46edb1e790e35d35a38b1a8f61a7ab82bddc52cd2c76e1fb71e54972c526746d
|
| 3 |
+
size 165
|