• Joined on 2024-05-31
tacit synced commits to refs/pull/425/merge at tacit/nanochat from mirror 2026-01-12 22:43:44 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/370/merge at tacit/nanochat from mirror 2026-01-12 22:43:43 +00:00
4610a838a1 record negative result on MTP
Compare 2 commits »
tacit synced commits to refs/pull/151/merge at tacit/nanochat from mirror 2026-01-12 22:43:43 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/396/merge at tacit/nanochat from mirror 2026-01-12 22:43:43 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/141/merge at tacit/nanochat from mirror 2026-01-12 22:43:43 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/328/merge at tacit/nanochat from mirror 2026-01-12 22:43:43 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
b33e394528 oops actually make SSSL the default window pattern
Compare 19 commits »
tacit synced commits to refs/pull/324/merge at tacit/nanochat from mirror 2026-01-12 22:43:43 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/59/merge at tacit/nanochat from mirror 2026-01-12 14:33:50 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/433/merge at tacit/nanochat from mirror 2026-01-12 14:33:50 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/429/merge at tacit/nanochat from mirror 2026-01-12 14:33:49 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/312/merge at tacit/nanochat from mirror 2026-01-12 14:33:49 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
b33e394528 oops actually make SSSL the default window pattern
Compare 22 commits »
tacit synced commits to refs/pull/400/merge at tacit/nanochat from mirror 2026-01-12 14:33:49 +00:00
4610a838a1 record negative result on MTP
Compare 2 commits »
tacit synced commits to refs/pull/399/merge at tacit/nanochat from mirror 2026-01-12 14:33:49 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
b33e394528 oops actually make SSSL the default window pattern
Compare 10 commits »
tacit synced commits to refs/pull/296/merge at tacit/nanochat from mirror 2026-01-12 14:33:48 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
Compare 4 commits »
tacit synced commits to refs/pull/204/merge at tacit/nanochat from mirror 2026-01-12 14:33:48 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
b33e394528 oops actually make SSSL the default window pattern
Compare 10 commits »
tacit synced commits to refs/pull/311/merge at tacit/nanochat from mirror 2026-01-12 14:33:48 +00:00
4610a838a1 record negative result on MTP
21608ec51e allow base_loss to report the loss of any arbitrary huggingface model similar to base_eval. had to change dataloader to be a lot better and just take tokenizer, not load the nanochat one. much better this way anyway
aa95fb2e03 make miniseries more generic and easier to run and less hard coded
b33e394528 oops actually make SSSL the default window pattern
Compare 22 commits »
tacit synced commits to refs/pull/59/merge at tacit/nanochat from mirror 2026-01-12 06:23:55 +00:00
b33e394528 oops actually make SSSL the default window pattern
fbc1484e8c add alternating window size patterns for the GPT layers, following GPT-3. Experimented a bit and found the pattern SSSL to work well - 3 short, 1 long alternating. This is now the new default and the plots look quite a bit better on flops vs. bpb
2ff7d51252 integrate Flash Attention 3. +9% tok_per_sec for d12 with ctx even as low as 2048 out of the box nice. also, ready to tune windows huge
201d705957 recover the ability to load old checkpoints by patching the lambdas if they don't exist in checkpoints
Compare 7 commits »
tacit synced commits to refs/pull/85/merge at tacit/nanochat from mirror 2026-01-12 06:23:55 +00:00
b33e394528 oops actually make SSSL the default window pattern
fbc1484e8c add alternating window size patterns for the GPT layers, following GPT-3. Experimented a bit and found the pattern SSSL to work well - 3 short, 1 long alternating. This is now the new default and the plots look quite a bit better on flops vs. bpb
2ff7d51252 integrate Flash Attention 3. +9% tok_per_sec for d12 with ctx even as low as 2048 out of the box nice. also, ready to tune windows huge
201d705957 recover the ability to load old checkpoints by patching the lambdas if they don't exist in checkpoints
Compare 7 commits »
tacit synced commits to refs/pull/93/merge at tacit/nanochat from mirror 2026-01-12 06:23:55 +00:00
b33e394528 oops actually make SSSL the default window pattern
fbc1484e8c add alternating window size patterns for the GPT layers, following GPT-3. Experimented a bit and found the pattern SSSL to work well - 3 short, 1 long alternating. This is now the new default and the plots look quite a bit better on flops vs. bpb
2ff7d51252 integrate Flash Attention 3. +9% tok_per_sec for d12 with ctx even as low as 2048 out of the box nice. also, ready to tune windows huge
201d705957 recover the ability to load old checkpoints by patching the lambdas if they don't exist in checkpoints
Compare 7 commits »
tacit synced commits to refs/pull/433/merge at tacit/nanochat from mirror 2026-01-12 06:23:54 +00:00
b33e394528 oops actually make SSSL the default window pattern
fbc1484e8c add alternating window size patterns for the GPT layers, following GPT-3. Experimented a bit and found the pattern SSSL to work well - 3 short, 1 long alternating. This is now the new default and the plots look quite a bit better on flops vs. bpb
Compare 3 commits »