vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

llama : new sampling algorithms (#1126)

Browse source 
* Sample interface, new samplers.

New samplers:
- locally typical sampling
- tail free sampling
- frequency and presence penalty
- mirostat

Ignore EOS fix: -inf should be used.

* mirostat

* Added --logit-bias and --no-penalize-nl, removed std::span

* Use C++11, clarify llama API documentation, rename Mirostat parameters to --mirostat_lr and --mirostat_ent, add temperature sampling for Mirostat, simplify Mirostat sampling API parameters (removed N and *k)

Use C++11, clarify llama API documentation, rename Mirostat parameters to --mirostat_lr and --mirostat_ent, add temperature sampling for Mirostat, simplify Mirostat sampling API parameters (removed N and *k)

* Save and load example adjust

* Tests

* Windows build fix

* Windows test fix
This commit is contained in:
Ivan Stepanov 2023-04-29 08:34:41 +03:00 • committed by GitHub
parent 7fc50c051a
commit dd7eff57d8
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
8 changed files with 812 additions and 160 deletions
Download patch file Download diff file Expand all files Collapse all files

1
tests/CMakeLists.txt
View file

@ -8,4 +8,5 @@ endfunction()
# llama_add_test(test-double-float.c) # SLOW
llama_add_test(test-quantize-fns.cpp)
llama_add_test(test-quantize-perf.cpp)
llama_add_test(test-sampling.cpp)
llama_add_test(test-tokenizer-0.cpp ${CMAKE_CURRENT_SOURCE_DIR}/../models/ggml-vocab.bin)