text-generation-inference
Enable multiple LoRa adapters
#2010
Merged

Enable multiple LoRa adapters #2010

drbh merged 41 commits into main from lora-internal
drbh
flozi00
drbh
flozi00
drbh
flozi00
drbh
drbh drbh force pushed from 091f2dce to d103264e 1 year ago
HuggingFaceDocBuilderDev
tgaddair
drbh drbh marked this pull request as ready for review 1 year ago
drbh
drbh drbh changed the title Lora internal Enable multiple LoRa adapters 1 year ago
drbh drbh requested a review from OlivierDehaene OlivierDehaene 1 year ago
drbh drbh requested a review from Narsil Narsil 1 year ago
drbh drbh requested a review from danieldk danieldk 1 year ago
flozi00
drbh feat: first draft load multiple lora
db3d8e65
drbh feat: load weights within layer and refactor lora pass
0a6ea7fb
drbh fix: refactor and reduce lora math
a046c303
drbh feat: baseline impl single request multi lora support
c6616312
drbh feat: prefer lorax implementation and port loading logic
8b50f4b7
drbh fix: prefer adapter_data and refactors
d5f21d57
drbh feat: perfer loraxs custom punica kernels and add mlp loras
8984ce6c
drbh fix: adjust batch for bgmv
ad088d51
drbh fix: adjust adapter_segments logic when in batch
c9273767
drbh fix: refactor and move changes to v3 proto
73eb2ae2
drbh fix: pass model_id for all flash causal lms
88bd5c2c
drbh fix: pass model_id for all causal and seq2seq lms
dc0f7655
drbh fix: add model_id to model test
9c45d349
drbh feat: add lora support to mistral and refactors
de56a81c
drbh feat: prefer model id in request
68399c1a
drbh fix: include rust code for adapter id
81707bfb
drbh feat: bump launcher and add new lora docs
43ec9dfe
drbh feat: support base model generation and refactors
611225f0
drbh fix: rename doc to retry ci build
a563a931
drbh feat: support if vlm models
91f40722
drbh fix: add adapter_data param and avoid missing layers
b1169273
drbh fix: add adapter_data param to phi and neox
1deb3725
drbh fix: update all models forwards to include adapter_data
101b95ad
drbh fix: add model_id to IdeficsCausalLM
ce40ad26
datavistics Update lora.md
1be1ebc4
datavistics Update lora.md
d6cf63ca
drbh drbh force pushed from 5a0ed2b3 to d6cf63ca 1 year ago
xiadingZ
drbh fix: add lora kernel to dockerfile, support running without kernels a…
aa88c4fd
drbh fix: avoid dockerfile conflict
06c3254c
drbh fix: merge 'main' into lora-internal to resolve conflicts
0e1c28ca
drbh Merge branch 'main' into lora-internal
1104885f
drbh Merge branch 'main' into lora-internal
224455f3
drbh
xiadingZ
danieldk
danieldk commented on 2024-06-19
drbh fix: refactors and adjust flash llama lora logic
4f1543d3
drbh fix: skip llama test due to CI issue (temp)
ce70fce9
drbh fix: skip llama test CI (temp) 2
c9e4526b
drbh fix: revert skips and prefer updated ci token for tests
a07b6129
danieldk
danieldk commented on 2024-06-20
danieldk
danieldk commented on 2024-06-21
danieldk
drbh fix: refactors and helpful comments
3c9b28ea
drbh fix: add noop in TensorParallelAdapterRowLinear too
c927cffb
drbh fix: refactor and move shard_lora_weights logic
f94f2b3e
drbh Merge branch 'main' into lora-internal
0d496baa
danieldk
danieldk dismissed these changes on 2024-06-25
drbh fix: exit early if no adapter_data
a2d821c4
drbh drbh dismissed their stale review via a2d821c4 1 year ago
drbh Merge branch 'main' into lora-internal
59575fe6
drbh
drbh drbh merged 04e1af94 into main 1 year ago
drbh drbh deleted the lora-internal branch 1 year ago
mhou7712

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone