| .. |
|
__init__.py
|
07aa2a492f
upstream: add option to specify tokenizer
|
hace 1 año |
|
block.py
|
ac82b67f75
feat: naive context shift and various QoL changes (#289)
|
hace 10 meses |
|
config.py
|
c41462cfcd
feat: exllamav2 quantization (#305)
|
hace 10 meses |
|
gguf.py
|
e42a78381a
feat: switch from pylint to ruff (#322)
|
hace 10 meses |
|
grammar.py
|
0527131e93
fix: grammar logits processor (#268)
|
hace 10 meses |
|
logger.py
|
e42a78381a
feat: switch from pylint to ruff (#322)
|
hace 10 meses |
|
logits_processor.py
|
3df36ee07d
fix: logit bias logitproc (#278)
|
hace 10 meses |
|
outputs.py
|
e42a78381a
feat: switch from pylint to ruff (#322)
|
hace 10 meses |
|
sampling_params.py
|
e42a78381a
feat: switch from pylint to ruff (#322)
|
hace 10 meses |
|
sequence.py
|
72659e5cad
separate prompt and genned tokens for enc-dec
|
hace 9 meses |
|
test_utils.py
|
641bb0f6e9
feat: add custom allreduce kernels (#224)
|
hace 11 meses |
|
utils.py
|
e53842bd5d
fix: cuda home detection for fp8 kv cache
|
hace 9 meses |
AlpinDale
72659e5cad
separate prompt and genned tokens for enc-dec