apunkt/doc-to-lora
1
0
Fork 0
mirror of https://github.com/SakanaAI/doc-to-lora.git synced 2026-07-23 17:01:04 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

bigger perceiver scripts

Browse source
This commit is contained in:
51616 2025-07-08 06:40:47 +00:00
parent 3f174b52f8
commit 63ecb73fdb
2 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

4
scripts/short_ctx/gemma_qa_short_ctx_perceiver.sh
View file

@ -18,8 +18,8 @@ uv run accelerate launch --main_process_port $port \
--gradient_accumulation_steps=2 \
--per_device_eval_batch_size=64 \
--target_modules=down_proj \
--num_blocks=4 \
--num_self_attn_per_block=3 \
--num_blocks=8 \
--num_self_attn_per_block=6 \
--n_latent_queries=208 \
--num_pre_head_layers=1 \
--lora_r=8 \

4
scripts/short_ctx/gemma_qa_short_ctx_perceiver_l2l.sh
View file

@ -18,8 +18,8 @@ uv run accelerate launch --main_process_port $port \
--gradient_accumulation_steps=4 \
--per_device_eval_batch_size=32 \
--target_modules=down_proj \
--num_blocks=4 \
--num_self_attn_per_block=3 \
--num_blocks=8 \
--num_self_attn_per_block=6 \
--ctx_encoder_type=per_layer_activations \
--n_latent_queries=8 \
--num_pre_head_layers=1 \