v0.2.5

What's Changed

• Add citation by @hugoabonizio in #81
• Faster startup using less memory in local mode by @hugoabonizio in #82

Full Changelog: v0.2.4...v0.2.5

v0.2.4

What's Changed

• Update MariTalk Local docs by @hugoabonizio in #52
• Add L4 GPU in the description by @rodrigo-f-nogueira in #73
• Use L4 GPU by default because it is easier to get one by @rodrigonogueira4 in #74
• Add hardware requirements by @hugoabonizio in #76
• Add more info about MariTalk Local by @rodrigo-f-nogueira in #77
• Improve installation/troubleshooting instructions by @rodrigo-f-nogueira in #75
• Updates the request example by @thiagolaitz in #78
• Improve local version starting by @hugoabonizio in #79

Full Changelog: v0.2.3...v0.2.4

v0.2.3

What's Changed

• add installation instruction by @rodrigo-f-nogueira in #63
• Delete outdated maritalk+langchain notebook by @rodrigo-f-nogueira in #65
• use new generate() method by @rodrigo-f-nogueira in #66
• Update llama-index link by @rodrigo-f-nogueira in #67
• Format commands so we can copy and paste them by @rodrigo-f-nogueira in #68
• Add missing model argument by @rodrigo-f-nogueira in #72
• Use full CUDA version by @hugoabonizio in #71

Full Changelog: v0.2.1...v0.2.3

v0.2.1

What's Changed

• Fix default model by @hugoabonizio in #58
• Update link to langchain documentation by @rodrigo-f-nogueira in #60
• Improve Streaming documentation and split API vs Local documentation by @rodrigo-f-nogueira in #61
• remove ssl dependency by @ZanezZephyrs in #59
• Add download and start commands by @hugoabonizio in #62

Full Changelog: v0.2.0...v0.2.1

v0.2.0

What's Changed

• Add link to langchain by @rodrigo-f-nogueira in #54
• Atualizações sabiá-2 by @rodrigo-f-nogueira in #55
• Stream by @thiagolaitz in #57

New Contributors

• @thiagolaitz made their first contribution in #57

Full Changelog: v0.1.10...v0.2.0

v0.1.10

What's Changed

• Remove unnecessary print by @rodrigo-f-nogueira in #37
• Add more explicit link to OCI guide by @rodrigo-f-nogueira in #36
• Melhorar descrição de verificação de meios de pagamento by @rodrigo-f-nogueira in #41
• Better name by @rodrigonogueira4 in #30
• Add GCP tutorial by @hugoabonizio in #25
• Add tokenizer example by @rodrigo-f-nogueira in #43
• Breaking Change: Return dict instead of string by @rodrigo-f-nogueira in #44
• Add model name by @rodrigo-f-nogueira in #46
• Improve documentation by @rodrigo-f-nogueira in #47
• Use full link paths so pdoc does not break links by @rodrigo-f-nogueira in #48
• Fix links by @rodrigo-f-nogueira in #49
• Fix examples folder link when rendered with pdoc by @rodrigo-f-nogueira in #50
• Add llama-index example by @rodrigo-f-nogueira in #51
• Rename tokenizer by @rodrigo-f-nogueira in #53

New Contributors

• @rodrigonogueira4 made their first contribution in #30

Full Changelog: v0.1.8...v0.1.10

v0.1.8

v0.1.8

0.1.7

v0.1.7

v0.1.7

0.1.5

Remove pycuda and rely on nvidia-smi output to detect the device comp…

…ute capability (#15)

0.1.4

Improve error handling on local deploy (#12)

* Improve error handling during the download process and version checking

* Show the error when the process fails to start

* Remove pycuda as global dependency and only rely on it in MariTalkLocal

* Add more documentation