Ollama v0.1.33

https://github.com/ollama/ollama/releases/tag/v0.1.33

New models:

Llama 3: a new model by Meta, and the most capable openly available LLM to date
Phi 3 Mini: a new 3.8B parameters, lightweight, state-of-the-art open model by Microsoft.
Moondream moondream is a small vision language model designed to run efficiently on edge devices.
Llama 3 Gradient 1048K: A Llama 3 fine-tune by Gradient to support up to a 1M token context window.
Dolphin Llama 3: The uncensored Dolphin model, trained by Eric Hartford and based on Llama 3 with a variety of instruction, conversational, and coding skills.
Qwen 110B: The first Qwen model over 100B parameters in size with outstanding performance in evaluations

What's Changed

Fixed issues where the model would not terminate, causing the API to hang.
Fixed a series of out of memory errors on Apple Silicon Macs
Fixed out of memory errors when running Mixtral architecture models

Experimental concurrency features

New concurrency features are coming soon to Ollama. They are available

OLLAMA_NUM_PARALLEL: Handle multiple requests simultaneously for a single model
OLLAMA_MAX_LOADED_MODELS: Load multiple models simultaneously

To enable these features, set the environment variables for ollama serve. For more info see this guide:

OLLAMA_NUM_PARALLEL=4 OLLAMA_MAX_LOADED_MODELS=4 ollama serve

New Contributors

@hmartinez82 made their first contribution in #3972
@Cephra made their first contribution in #4037
@arpitjain099 made their first contribution in #4007
@MarkWard0110 made their first contribution in #4031
@alwqx made their first contribution in #4073
@Sidxt made their first contribution in #3705
@ChengenH made their first contribution in #3789
@secondtruth made their first contribution in #3503
@reid41 made their first contribution in #3612
@ericcurtin made their first contribution in #3626
@JT2M0L3Y made their first contribution in #3633
@datvodinh made their first contribution in #3655
@MapleEve made their first contribution in #3817
@swuecho made their first contribution in #3810
@brycereitano made their first contribution in #3895
@bsdnet made their first contribution in #3889
@fyxtro made their first contribution in #3855
@natalyjazzviolin made their first contribution in #3962

Full Changelog: v0.1.32...v0.1.33

{
  "by": "tosh",
  "descendants": 0,
  "id": 40247977,
  "score": 2,
  "time": 1714746050,
  "title": "Ollama v0.1.33",
  "type": "story",
  "url": "https://github.com/ollama/ollama/releases/tag/v0.1.33"
}

{
  "author": "ollama",
  "date": null,
  "description": "New models: Llama 3: a new model by Meta, and the most capable openly available LLM to date\nPhi 3 Mini: a new 3.8B parameters, lightweight, state-of-the-art open model by Microsoft.\nMoondream moon…",
  "image": "https://opengraph.githubassets.com/2c8cc5a47f09a7f89e4b784e1e2558997edead372d779b2f81b468f197a2b641/ollama/ollama/releases/tag/v0.1.33",
  "logo": "https://logo.clearbit.com/github.com",
  "publisher": "GitHub",
  "title": "Release v0.1.33 · ollama/ollama",
  "url": "https://github.com/ollama/ollama/releases/tag/v0.1.33"
}

{
  "url": "https://github.com/ollama/ollama/releases/tag/v0.1.33",
  "title": "Release v0.1.33 · ollama/ollama",
  "description": "New models:\n\nLlama 3: a new model by Meta, and the most capable openly available LLM to date\nPhi 3 Mini: a new 3.8B parameters, lightweight, state-of-the-art open model by Microsoft.\nMoondream moon...",
  "links": [
    "https://github.com/ollama/ollama/releases/tag/v0.1.33"
  ],
  "image": "https://opengraph.githubassets.com/2c8cc5a47f09a7f89e4b784e1e2558997edead372d779b2f81b468f197a2b641/ollama/ollama/releases/tag/v0.1.33",
  "content": "<div><p><a target=\"_blank\" href=\"https://private-user-images.githubusercontent.com/3325447/326950213-8dc9c472-9d72-4b39-95ae-2c85ada375b9.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.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.zhVaay_YIuoADtqSDwAiog-F7ED1D__Pl2G3-a0CYfw\"><img src=\"https://private-user-images.githubusercontent.com/3325447/326950213-8dc9c472-9d72-4b39-95ae-2c85ada375b9.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.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.zhVaay_YIuoADtqSDwAiog-F7ED1D__Pl2G3-a0CYfw\" alt=\"Llama 3\" /></a></p>\n<h2>New models:</h2>\n<ul>\n<li><a target=\"_blank\" href=\"https://ollama.com/library/llama3\">Llama 3</a>: a new model by Meta, and the most capable openly available LLM to date</li>\n<li><a target=\"_blank\" href=\"https://ollama.com/library/phi3\">Phi 3 Mini</a>: a new 3.8B parameters, lightweight, state-of-the-art open model by Microsoft.</li>\n<li><a target=\"_blank\" href=\"https://ollama.com/library/moondream\">Moondream</a> moondream is a small vision language model designed to run efficiently on edge devices.</li>\n<li><a target=\"_blank\" href=\"https://ollama.com/library/llama3-gradient\">Llama 3 Gradient 1048K</a>: A Llama 3 fine-tune by Gradient to support up to a 1M token context window.</li>\n<li><a target=\"_blank\" href=\"https://ollama.com/library/dolphin-llama3\">Dolphin Llama 3</a>: The uncensored Dolphin model, trained by Eric Hartford and based on Llama 3 with a variety of instruction, conversational, and coding skills.</li>\n<li><a target=\"_blank\" href=\"https://ollama.com/library/qwen:110b\">Qwen 110B</a>: The first Qwen model over 100B parameters in size with outstanding performance in evaluations</li>\n</ul>\n<h2>What's Changed</h2>\n<ul>\n<li>Fixed issues where the model would not terminate, causing the API to hang.</li>\n<li>Fixed a series of out of memory errors on Apple Silicon Macs</li>\n<li>Fixed out of memory errors when running Mixtral architecture models</li>\n</ul>\n<h2>Experimental concurrency features</h2>\n<p>New concurrency features are coming soon to Ollama. They are available</p>\n<ul>\n<li><code>OLLAMA_NUM_PARALLEL</code>: Handle multiple requests simultaneously for a single model</li>\n<li><code>OLLAMA_MAX_LOADED_MODELS</code>: Load multiple models simultaneously</li>\n</ul>\n<p>To enable these features, set the environment variables for <code>ollama serve</code>. For more info see <a target=\"_blank\" href=\"https://github.com/ollama/ollama/blob/main/docs/faq.md#how-do-i-configure-ollama-server\">this guide</a>:</p>\n<div><pre><code>OLLAMA_NUM_PARALLEL=4 OLLAMA_MAX_LOADED_MODELS=4 ollama serve\n</code></pre></div>\n<h2>New Contributors</h2>\n<ul>\n<li><a target=\"_blank\" href=\"https://github.com/hmartinez82\">@hmartinez82</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3972\">#3972</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/Cephra\">@Cephra</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/4037\">#4037</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/arpitjain099\">@arpitjain099</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/4007\">#4007</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/MarkWard0110\">@MarkWard0110</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/4031\">#4031</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/alwqx\">@alwqx</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/4073\">#4073</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/Sidxt\">@Sidxt</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3705\">#3705</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/ChengenH\">@ChengenH</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3789\">#3789</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/secondtruth\">@secondtruth</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3503\">#3503</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/reid41\">@reid41</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3612\">#3612</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/ericcurtin\">@ericcurtin</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3626\">#3626</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/JT2M0L3Y\">@JT2M0L3Y</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3633\">#3633</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/datvodinh\">@datvodinh</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3655\">#3655</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/MapleEve\">@MapleEve</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3817\">#3817</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/swuecho\">@swuecho</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3810\">#3810</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/brycereitano\">@brycereitano</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3895\">#3895</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/bsdnet\">@bsdnet</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3889\">#3889</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/fyxtro\">@fyxtro</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3855\">#3855</a></li>\n<li><a target=\"_blank\" href=\"https://github.com/natalyjazzviolin\">@natalyjazzviolin</a> made their first contribution in <a target=\"_blank\" href=\"https://github.com/ollama/ollama/pull/3962\">#3962</a></li>\n</ul>\n<p><strong>Full Changelog</strong>: <a target=\"_blank\" href=\"https://github.com/ollama/ollama/compare/v0.1.32...v0.1.33\">v0.1.32...v0.1.33</a></p></div>",
  "author": "",
  "favicon": "https://github.githubassets.com/favicons/favicon.svg",
  "source": "github.com",
  "published": "",
  "ttr": 65,
  "type": "object"
}