Accio-Lab
/

occamy-1.0-MTP

speculative-decoding

Model card Files Files and versions

1.69 GB

Ctrl+K

Ctrl+K

1 contributor

History: 5 commits

Eang's picture

Link four MLX precisions and add local API usage where applicable

808a5e3 verified 22 days ago

runtime
Add validated native MTP3 runtime and H200 results about 1 month ago
validation
Add validated native MTP3 runtime and H200 results about 1 month ago
.gitattributes

1.52 kB
Release Occamy 1.0 MTP about 1 month ago
LICENSE

11.4 kB
Release Occamy 1.0 MTP about 1 month ago
NATIVE-MTP3.md

3.78 kB
Add validated native MTP3 runtime and H200 results about 1 month ago
PUBLICATION-GATE.json

467 Bytes
Release Occamy 1.0 MTP about 1 month ago
README.md

7.35 kB
Link four MLX precisions and add local API usage where applicable 22 days ago
SHA256SUMS

1.22 kB
Release Occamy 1.0 MTP about 1 month ago
TECHNICAL-DETAILS.md

5.84 kB
Release Occamy 1.0 MTP about 1 month ago
TRAINING.json

1.81 kB
Release Occamy 1.0 MTP about 1 month ago
VALIDATION.json

5.02 kB
Release Occamy 1.0 MTP about 1 month ago
VLLM-COMMUNITY.md

2.35 kB
Preserve BF16 MTP head in NVFP4 assembly and document community vLLM setup about 1 month ago
assemble_head.py

3.41 kB
Preserve BF16 MTP head in NVFP4 assembly and document community vLLM setup about 1 month ago
evaluation_cases.json

4.62 kB
Release Occamy 1.0 MTP about 1 month ago
mtp-trained.safetensors

1.69 GB
xet

Release Occamy 1.0 MTP about 1 month ago
mtp_head_config.json

1.04 kB
Release Occamy 1.0 MTP about 1 month ago
per-request-results.json

144 kB
Release Occamy 1.0 MTP about 1 month ago
test_assemble_head.py

2.9 kB
Preserve BF16 MTP head in NVFP4 assembly and document community vLLM setup about 1 month ago