ElizaWszola
b00b33d77e
[Model][Quantization] HQQ support through Marlin kernel expansion ( #9766 )
...
Signed-off-by: ElizaWszola <eliza@neuralmagic.com >
2024-11-19 13:31:12 -08:00
Jee Jee Li
7eb719df13
[Bugfix]Fix Phi-3 BNB online quantization ( #10417 )
...
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com >
2024-11-19 03:21:42 +00:00
Yan Ma
6b2d25efc7
[Hardware][XPU] AWQ/GPTQ support for xpu backend ( #10107 )
...
Signed-off-by: yan ma <yan.ma@intel.com >
2024-11-18 11:18:05 -07:00
Li, Jiang
ca77dd7a44
[Hardware][CPU] Support AWQ for CPU backend ( #7515 )
2024-10-09 10:28:08 -06:00
chenqianfzh
2f4117c38e
support bitsandbytes quantization with more models ( #9148 )
2024-10-08 19:52:19 -06:00
Isotr0py
f19da64871
[Core] Refactor GGUF parameters packing and forwarding ( #8859 )
2024-10-07 10:01:46 +00:00
chenqianfzh
9855b99502
[Feature][kernel] tensor parallelism with bitsandbytes quantization ( #8434 )
2024-09-17 08:09:12 -07:00
Pavani Majety
efcf946a15
[Hardware][NV] Add support for ModelOpt static scaling checkpoints. ( #6112 )
2024-09-11 00:38:40 -04:00
Dipika Sikka and Michael Goin
e16fa99a6a
[Misc] Update fbgemmfp8 to use vLLMParameters ( #7972 )
...
Co-authored-by: Michael Goin <michael@neuralmagic.com >
2024-09-03 20:12:41 -06:00
Dipika Sikka
2188a60c7e
[Misc] Update GPTQ to use vLLMParameters ( #7976 )
2024-09-03 17:21:44 -04:00
chenqianfzh
4664ceaad6
support bitsandbytes 8-bit and FP4 quantized models ( #7445 )
2024-08-29 19:09:08 -04:00
Dipika Sikka
86a677de42
[misc] update tpu int8 to use new vLLM Parameters ( #7973 )
2024-08-29 16:46:55 -04:00
Dipika Sikka
015e6cc252
[Misc] Update compressed tensors lifecycle to remove prefix from create_weights ( #7825 )
2024-08-26 18:09:34 -06:00
Dipika Sikka and Michael Goin
dd9857f5fa
[Misc] Update gptq_marlin_24 to use vLLMParameters ( #7762 )
...
Co-authored-by: Michael Goin <michael@neuralmagic.com >
2024-08-26 17:44:54 -04:00
Dipika Sikka
665304092d
[Misc] Update qqq to use vLLMParameters ( #7805 )
2024-08-26 13:16:15 -06:00
Dipika Sikka
f1df5dbfd6
[Misc] Update marlin to use vLLMParameters ( #7803 )
2024-08-23 14:30:52 -04:00
Dipika Sikka
955b5191c9
[Misc] update fp8 to use vLLMParameter ( #7437 )
2024-08-22 08:36:18 -04:00
Isotr0py
12e1c65bc9
[Model] Add AWQ quantization support for InternVL2 model ( #7187 )
2024-08-20 23:18:57 -07:00
Isotr0py
7601cb044d
[Core] Support tensor parallelism for GGUF quantization ( #7520 )
2024-08-19 17:30:14 -04:00
Dipika Sikka
b1e5afc3e7
[Misc] Update awq and awq_marlin to use vLLMParameters ( #7422 )
2024-08-13 17:08:20 -04:00
Dipika Sikka
fb377d7e74
[Misc] Update gptq_marlin to use new vLLMParameters ( #7281 )
2024-08-13 14:30:11 -04:00
Dipika Sikka
5c6c54d67a
[Bugfix] Fix PerTensorScaleParameter weight loading for fused models ( #7376 )
2024-08-09 21:23:46 +00:00
Dipika Sikka
0f7052bc7e
[Misc] Refactor linear layer weight loading; introduce BasevLLMParameter and weight_loader_v2 ( #5874 )
2024-08-07 09:17:58 -07:00
Isotr0py and Michael Goin
360bd67cf0
[Core] Support loading GGUF model ( #5191 )
...
Co-authored-by: Michael Goin <michael@neuralmagic.com >
2024-08-05 17:54:23 -06:00
QQSong
062a1d0fab
Fix ReplicatedLinear weight loading ( #6793 )
2024-07-25 19:24:58 -07:00
Robert Shaw
683e3cb9c4
[ Misc ] fbgemm checkpoints ( #6559 )
2024-07-20 09:36:57 -07:00
Thomas Parnell
a5314e8698
[Model] RowParallelLinear: pass bias to quant_method.apply ( #6327 )
...
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com >
2024-07-19 07:15:22 -06:00
Robert Shaw
dbe5588554
[ Misc ] non-uniform quantization via compressed-tensors for Llama ( #6515 )
2024-07-18 22:39:18 -04:00
Michael Goin
978aed5300
[Kernel][Attention] Separate Attention.kv_scale into k_scale and v_scale ( #6081 )
2024-07-16 15:31:32 -07:00
Robert Shaw
6047187cd8
[ Misc ] Remove separate bias add ( #6353 )
2024-07-12 05:06:09 +00:00
daquexian
99ded1e1c4
[Doc] Remove comments incorrectly copied from another project ( #6286 )
2024-07-10 17:05:26 -07:00
Baoyuan Qi and Robert Shaw
d3a245138a
[Bugfix]fix and needs_scalar_to_array logic check ( #6238 )
...
Co-authored-by: Robert Shaw <114415538+robertgshaw2-neuralmagic@users.noreply.github.com >
2024-07-09 23:43:24 +00:00
Robert Shaw and Robert Shaw
af9ad46fca
[ Misc ] Refactor w8a8 to use process_weights_after_load (Simplify Weight Loading) ( #5940 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
2024-06-30 23:06:27 +00:00
Robert Shaw and Robert Shaw
2cd402e169
[ Bugfix ] Enabling Loading Models With Fused QKV/MLP on Disk with FP8 ( #5921 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
2024-06-28 18:43:49 +00:00
Robert Shaw and Robert Shaw
b185230744
[ Misc ] Remove fp8_shard_indexer from Col/Row Parallel Linear (Simplify Weight Loading) ( #5928 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
2024-06-28 13:49:57 -04:00
Dipika Sikka
95db455e7f
[Misc] Add channel-wise quantization support for w8a8 dynamic per token activation quantization ( #5542 )
2024-06-18 12:45:05 -04:00
Cyrus Leung
0e9164b40a
[mypy] Enable type checking for test directory ( #5017 )
2024-06-15 04:45:31 +00:00
chenqianfzh
b9c0605a8e
[Feature][Kernel] Support bitsandbytes quantization and QLoRA ( #4776 )
2024-06-01 14:51:10 -06:00
a1242324c9
[Kernel] Initial Activation Quantization Support ( #4525 )
...
Co-authored-by: Varun Sundar Rabindranath <varunsundar08@gmail.com >
Co-authored-by: Varun Sundar Rabindranath <varun@neuralmagic.com >
2024-05-23 21:29:18 +00:00
Jee Li
d6f4bd7cdd
[Misc]Add customized information for models ( #4132 )
2024-04-30 21:18:14 -07:00
111815d482
[Kernel] Support Fp8 Checkpoints (Dynamic + Static) ( #4332 )
...
Co-authored-by: Philipp Moritz <pcmoritz@gmail.com >
Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu >
Co-authored-by: mgoin <michael@neuralmagic.com >
Co-authored-by: Tyler Michael Smith <tyler@neuralmagic.com >
Co-authored-by: Cody Yu <hao.yu.cody@gmail.com >
2024-04-30 21:46:12 +00:00
SangBin Cho
df29793dc7
[mypy][5/N] Support all typing on model executor ( #4427 )
2024-04-28 19:01:26 -07:00
Cody Yu
a62aaf1df5
[Misc][Refactor] Generalize linear_method to be quant_method ( #4373 )
2024-04-26 16:41:14 -04:00
Robert Shaw
79a268c4ab
[BUG] fixed fp8 conflict with aqlm ( #4307 )
...
Fixes fp8 iterface which broke in AQLM merge.
2024-04-23 18:26:33 -07:00
James Fleming and mgoin
2b7949c1c2
AQLM CUDA support ( #3287 )
...
Co-authored-by: mgoin <michael@neuralmagic.com >
2024-04-23 13:59:33 -04:00
Cody Yu
a22cdea371
[Kernel][FP8] Initial support with dynamic per-tensor scaling ( #4118 )
...
Provide an initial support to FP8 computation. This PR is inspired by HuggingFace TGI: huggingface/text-generation-inference#1726
This feature can be enabled with --quantization fp8 or -q fp8 when launching an engine.
Algorithm:
We still load a model checkpoint in FP16/BF16. After the weights are loaded, Fp8LinearMethod calculates the per-tensor scaling factor of weights and quantizes the weights accordingly. The scaling factor will then be stored for future use. Meanwhile, the per-tensor scaling factor for activations is calculated in every forward pass.
Initial Results:
Currently tested Mistral-7B on 1xH100. With prompt length ~5 and decoding length 128:
BF16: 1.47s
FP8: 1.66s
I'll try to use larger models and try to find more performance bottleneck. Meanwhile, you're welcome to try this code.
2024-04-20 04:28:57 +00:00
Antoni Baum
a10d3056da
[Core] Set linear_weights directly on the layer ( #3977 )
2024-04-11 16:35:51 -04:00
youkaichao
63e7176f26
[Core][Refactor] move parallel_utils into vllm/distributed ( #3950 )
...
[WIP][Core][Refactor] move vllm/model_executor/parallel_utils into vllm/distributed and vllm/device_communicators (#3950 )
2024-04-10 15:33:30 -07:00
SangBin Cho
01bfb22b41
[CI] Try introducing isort. ( #3495 )
2024-03-25 07:59:47 -07:00
Hui Liu
ba8dc958a3
[Minor] Fix bias in if to remove ambiguity ( #3259 )
2024-03-13 09:16:55 -07:00
ElizaWszola