<feed xmlns='http://www.w3.org/2005/Atom'>
<title>ffwd/pplx_embed_mlx.c, branch master</title>
<subtitle>Fastforward embedding inference engine</subtitle>
<id>https://git.onurgunduz.com/ffwd/atom/pplx_embed_mlx.c?h=master</id>
<link rel='self' href='https://git.onurgunduz.com/ffwd/atom/pplx_embed_mlx.c?h=master'/>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/'/>
<updated>2026-06-14T06:27:03Z</updated>
<entry>
<title>gitignore: track new binary and library names</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-10T06:24:39Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=2c37727e653d2340d8977dd5db4addc767dd81ea'/>
<id>urn:sha1:2c37727e653d2340d8977dd5db4addc767dd81ea</id>
<content type='text'>
</content>
</entry>
<entry>
<title>kernels: drop asr_ prefix naming</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-10T06:02:03Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=ec417bcb52f2aa8e24aeec312e8ddbf3a0798056'/>
<id>urn:sha1:ec417bcb52f2aa8e24aeec312e8ddbf3a0798056</id>
<content type='text'>
</content>
</entry>
<entry>
<title>mlx: mixed-precision BF16 inference (F32 residual, BF16 weights)</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-08T23:52:28Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=d1e003ecae35aa10f89919e62b1dafbcd48aa07b'/>
<id>urn:sha1:d1e003ecae35aa10f89919e62b1dafbcd48aa07b</id>
<content type='text'>
</content>
</entry>
<entry>
<title>mlx: split ragged contextual batches</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-07T03:06:19Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=1c708cfbfab4361861b3ea5600af1d2bd8b7daf2'/>
<id>urn:sha1:1c708cfbfab4361861b3ea5600af1d2bd8b7daf2</id>
<content type='text'>
</content>
</entry>
<entry>
<title>mlx: tune ragged grouping for 4B</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-04T19:10:14Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=8cfad91097dcdcfd2b2a7c6f65cbe0eba58d4bf5'/>
<id>urn:sha1:8cfad91097dcdcfd2b2a7c6f65cbe0eba58d4bf5</id>
<content type='text'>
</content>
</entry>
<entry>
<title>mlx: prepare contiguous linear weights</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-04T18:18:19Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=cdbbd670fb1917edd625ff1c19f7a28f2852a930'/>
<id>urn:sha1:cdbbd670fb1917edd625ff1c19f7a28f2852a930</id>
<content type='text'>
</content>
</entry>
<entry>
<title>mlx: optimize ragged embedding batches</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-04T17:11:19Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=a89fc2041de88b2b6fb326b3031a602333068052'/>
<id>urn:sha1:a89fc2041de88b2b6fb326b3031a602333068052</id>
<content type='text'>
</content>
</entry>
<entry>
<title>chore: address static analyzer warnings</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-04T16:46:46Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=d08f68523ea9a39af8176703e48fd47a987893b5'/>
<id>urn:sha1:d08f68523ea9a39af8176703e48fd47a987893b5</id>
<content type='text'>
</content>
</entry>
<entry>
<title>mlx: use prefix pooling for f32 contextual spans</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-03T17:57:08Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=3605a86f42c84d5c0e8600180d30246588dcbf8e'/>
<id>urn:sha1:3605a86f42c84d5c0e8600180d30246588dcbf8e</id>
<content type='text'>
</content>
</entry>
<entry>
<title>mlx: cache transposed linear weights</title>
<updated>2026-06-14T06:27:03Z</updated>
<author>
<name>Onur Gündüz</name>
<email>ogunduz@gmail.com</email>
</author>
<published>2026-06-03T17:49:22Z</published>
<link rel='alternate' type='text/html' href='https://git.onurgunduz.com/ffwd/commit/?id=bd914f876842d5169616fa4eb257c44b55d4bf8b'/>
<id>urn:sha1:bd914f876842d5169616fa4eb257c44b55d4bf8b</id>
<content type='text'>
</content>
</entry>
</feed>
