Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Evo 2 more comprehensive notebook and 40B configs #203

Merged
merged 3 commits into from
Feb 26, 2025
Merged

Evo 2 more comprehensive notebook and 40B configs #203

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
a4776c6
Add comparison notebook between Helix-mRNA and Evo 2
mattwoodx Feb 26, 2025
88e679f
Add 40B configs for Evo 2
mattwoodx Feb 26, 2025
ffcc330
Update notebook
mattwoodx Feb 26, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1,723 changes: 1,362 additions & 361 deletions examples/notebooks/Evo-2.ipynb
View file
Open in desktop

Large diffs are not rendered by default.

16 changes: 2 additions & 14 deletions helical/models/evo_2/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,25 +4,13 @@

**Model Name:** Evo 2

**Model Versions:** 1B, 7B and 40B (Currently only the 1B and 7B models are available and the 40B coming soon)
**Model Versions:** 1B, 7B and 40B

**Model Description:** Evo 2 is a next-generation genomic model that integrates DNA, RNA, and protein data across all domains of life. It leverages the StripedHyena 2 architecture, combining convolutional, linear attention, and state-space models to efficiently process long sequences and capture complex biological patterns. Evo 2 is trained on a vast dataset encompassing trillions of nucleotides from eukaryotic and prokaryotic genomes, enabling broad cross-species applications and insights into human diseases, agriculture, and environmental science.

## Model Developers

**Arc Institute**

**Stanford University**

**NVIDIA**

**Liquid AI**

**University of California, Berkeley**

**Goodfire**

**Columbia University**
Arc Institute, Stanford University, NVIDIA, Liquid AI, University of California, Berkeley, Goodfire, Columbia University

**Contact Information:**

Expand Down
116 changes: 114 additions & 2 deletions helical/models/evo_2/evo_2_config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,14 +94,12 @@ def __init__(
"hcm_layer_idxs": [1, 5, 8, 12, 15, 19, 22, 26, 29],
"hcs_layer_idxs": [0, 4, 7, 11, 14, 18, 21, 25, 28],
"attn_layer_idxs": [3, 10, 17, 24, 31],

"hcm_filter_length": 128,
"hcl_filter_groups": 4096,
"hcm_filter_groups": 256,
"hcs_filter_groups": 256,
"hcs_filter_length": 7,
"num_layers": 32,

"short_filter_length": 3,
"num_attention_heads": 32,
"short_filter_bias": False,
Expand Down Expand Up @@ -205,6 +203,120 @@ def __init__(
"mlp_activation": "gelu",
"print_activations": False,
},
"evo2-40b-base": {
"model_name": "evo2_40b_base",
"model_hf_name": "arcinstitute/evo2_40b_base",
"default_embedding_layer": "blocks.49.mlp.l3",
"vocab_size": 512,
"hidden_size": 8192,
"num_filters": 8192,
"hcl_layer_idxs": [2, 6, 9, 13, 16, 20, 23, 27, 30, 34, 38, 41, 45, 48],
"hcm_layer_idxs": [1, 5, 8, 12, 15, 19, 22, 26, 29, 33, 37, 40, 44, 47],
"hcs_layer_idxs": [0, 4, 7, 11, 14, 18, 21, 25, 28, 32, 36, 39, 43, 46],
"attn_layer_idxs": [3, 10, 17, 24, 31, 35, 42, 49],
"hcm_filter_length": 128,
"hcl_filter_groups": 8192,
"hcm_filter_groups": 512,
"hcs_filter_groups": 512,
"hcs_filter_length": 7,
"num_layers": 50,
"short_filter_length": 3,
"num_attention_heads": 64,
"short_filter_bias": False,
"mlp_init_method": "torch.nn.init.zeros_",
"mlp_output_init_method": "torch.nn.init.zeros_",
"eps": 0.000001,
"state_size": 16,
"rotary_emb_base": 1000000,
"make_vocab_size_divisible_by": 8,
"inner_size_multiple_of": 128,
"inner_mlp_size": 21888,
"log_intermediate_values": False,
"proj_groups": 1,
"hyena_filter_groups": 1,
"column_split_hyena": False,
"column_split": True,
"interleave": True,
"evo2_style_activations": True,
"use_fp8_input_projections": True,
"model_parallel_size": 1,
"pipe_parallel_size": 1,
"tie_embeddings": True,
"mha_out_proj_bias": True,
"hyena_out_proj_bias": True,
"hyena_flip_x1x2": False,
"qkv_proj_bias": False,
"max_seqlen": 8192,
"max_batch_size": 1,
"final_norm": True,
"use_flash_attn": True,
"use_flash_rmsnorm": False,
"use_flash_depthwise": False,
"use_flashfft": False,
"use_laughing_hyena": False,
"inference_mode": True,
"prefill_style": "fft",
"mlp_activation": "gelu",
"print_activations": False,
},
"evo2-40b": {
"model_name": "evo2_40b",
"model_hf_name": "arcinstitute/evo2_40b",
"default_embedding_layer": "blocks.49.mlp.l3",
"vocab_size": 512,
"hidden_size": 8192,
"num_filters": 8192,
"hcl_layer_idxs": [2, 6, 9, 13, 16, 20, 23, 27, 30, 34, 38, 41, 45, 48],
"hcm_layer_idxs": [1, 5, 8, 12, 15, 19, 22, 26, 29, 33, 37, 40, 44, 47],
"hcs_layer_idxs": [0, 4, 7, 11, 14, 18, 21, 25, 28, 32, 36, 39, 43, 46],
"attn_layer_idxs": [3, 10, 17, 24, 31, 35, 42, 49],
"hcm_filter_length": 128,
"hcl_filter_groups": 8192,
"hcm_filter_groups": 512,
"hcs_filter_groups": 512,
"hcs_filter_length": 7,
"num_layers": 50,
"short_filter_length": 3,
"num_attention_heads": 64,
"short_filter_bias": False,
"mlp_init_method": "torch.nn.init.zeros_",
"mlp_output_init_method": "torch.nn.init.zeros_",
"eps": 0.000001,
"state_size": 16,
"rotary_emb_base": 100000000000,
"rotary_emb_scaling_factor": 128,
"use_interpolated_rotary_pos_emb": True,
"make_vocab_size_divisible_by": 8,
"inner_size_multiple_of": 128, # force GLU inner_size to be a multiple of
"inner_mlp_size": 22528,
"log_intermediate_values": False,
"proj_groups": 1,
"hyena_filter_groups": 1,
"column_split_hyena": False,
"column_split": True,
"interleave": True,
"evo2_style_activations": True,
"use_fp8_input_projections": True,
"model_parallel_size": 1,
"pipe_parallel_size": 1,
"tie_embeddings": True,
"mha_out_proj_bias": True,
"hyena_out_proj_bias": True,
"hyena_flip_x1x2": False,
"qkv_proj_bias": False,
"max_seqlen": 1048576,
"max_batch_size": 1,
"final_norm": True,
"use_flash_attn": True,
"use_flash_rmsnorm": False,
"use_flash_depthwise": False,
"use_flashfft": False,
"use_laughing_hyena": False,
"inference_mode": True,
"prefill_style": "fft",
"mlp_activation": "gelu",
"print_activations": False,
},
}
if model_name not in self.model_map:
raise ValueError(
Expand Down