Text Generation
NInfer
qwen.83
qwen3_8
blackwell
sm120a
mtp
speculative-decoding
indexed88
vocabulary-prune
coding
agentic
nvfp4
conversational
Instructions to use Schestex/ThinkingCap-Qwen3.8-27B-NInfer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NInfer
How to use Schestex/ThinkingCap-Qwen3.8-27B-NInfer with NInfer:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Upload folder using huggingface_hub
Browse files- provenance/build-meta.env +7 -0
- provenance/core-files.sha256 +2 -0
- provenance/git-diff.sha256 +1 -0
- provenance/git-status.txt +14 -0
- provenance/run.txt +12 -0
provenance/build-meta.env
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
BUILD_SECONDS=325
|
| 2 |
+
BASE_BYTES=17106573056
|
| 3 |
+
D48_BYTES=16705529348
|
| 4 |
+
ARTIFACT_BYTES=16575798788
|
| 5 |
+
SAVING_VS_BASE=530774268
|
| 6 |
+
SAVING_VS_D48=129730560
|
| 7 |
+
NVFP4_JOBS=92
|
provenance/core-files.sha256
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
bc9a5f4098bddea58651cedaf5866def6f8097d1ee1d32e3c61a455b931eb5f7 /home/ai/models/thinkingcap-d48-o44-perf-2026-09-27/override-d48-o44.py
|
| 2 |
+
73dc1d8e32cde31946e34e0badd378c19f442f37749575e80b11428b4f0e2f77 /home/ai/models/thinkingcap-d48-o44-perf-2026-09-27/nvfp4_encoder.py
|
provenance/git-diff.sha256
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
2e526a1eb2333992c146fde7324d7f615dc3e4a8dae1bc01f54357aaeed2befa /home/ai/models/thinkingcap-d48-o44-perf-2026-09-27/provenance/git-diff.patch
|
provenance/git-status.txt
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
M src/ops/dynamic_grouped_conv/sources.cmake
|
| 2 |
+
M src/ops/linear/q6/q6_dispatch.cpp
|
| 3 |
+
M src/ops/linear/q6/q6_shapes.h
|
| 4 |
+
M src/ops/linear/q6/sources.cmake
|
| 5 |
+
M src/ops/wrapper/dynamic_grouped_conv.cpp
|
| 6 |
+
M src/ops/wrapper/linear_add.cpp
|
| 7 |
+
?? DFLASH.md
|
| 8 |
+
?? Dockerfile.convert
|
| 9 |
+
?? NVFP4
|
| 10 |
+
?? compressed-tensors-style
|
| 11 |
+
?? src/ops/dynamic_grouped_conv/nvfp4/
|
| 12 |
+
?? src/ops/linear/q6/shapes/n5120_k17408.cpp
|
| 13 |
+
?? src/ops/linear/q6/shapes/n5120_k6144.cpp
|
| 14 |
+
?? tools/chat_templates/qwen3_8_sharp.jinja
|
provenance/run.txt
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
run=2026-09-27T07:17:02+00:00
|
| 2 |
+
repo=/home/ai/ninfer-v3-sharp-test
|
| 3 |
+
git_head=9e163eee4b8acec21ab0ac765107b6a3f287b217
|
| 4 |
+
base=/home/ai/models/thinkingcap-ninfer/thinkingcap-qwen3.8-27b.ninfer
|
| 5 |
+
base_sha256=42762ef3ddd7f3f2d5b53b7ccbec25882a99b759e04e8d814cbde899202a0e1c
|
| 6 |
+
d48=/home/ai/models/thinkingcap-down-sweep-2026-09-23/groups/d48/thinkingcap-d48.ninfer
|
| 7 |
+
d48_sha256=fdc304dca679849512e52c075194185193788ccaaf009148a9116f30eac69d36
|
| 8 |
+
convert_image=ninfer:a2-convert
|
| 9 |
+
bench_image=ninfer:a2-bench
|
| 10 |
+
candidate=D48+O44
|
| 11 |
+
quality_valid=NO_PERFORMANCE_DIAGNOSTIC
|
| 12 |
+
benchmark_reference=historical_d48_report_no_rerun
|