Add model files for Alpamayo R1 10b (#1)
Browse files- Add model files from AR1-test (1834af8ee83df0f3690bfae4e9b607fd4e54a383)
- Update config.json (fa32ec79664ba72d56f478e7d7870b67c21697e6)
Co-authored-by: Yu Wang <[email protected]>
- config.json +134 -0
- model-00001-of-00005.safetensors +3 -0
- model-00002-of-00005.safetensors +3 -0
- model-00003-of-00005.safetensors +3 -0
- model-00004-of-00005.safetensors +3 -0
- model-00005-of-00005.safetensors +3 -0
- model.safetensors.index.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,134 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"action_in_proj_cfg": {
|
| 3 |
+
"_target_": "alpamayo_r1.models.action_in_proj.PerWaypointActionInProjV2",
|
| 4 |
+
"hidden_size": 512,
|
| 5 |
+
"max_freq": 100.0,
|
| 6 |
+
"num_enc_layers": 2,
|
| 7 |
+
"num_fourier_feats": 20
|
| 8 |
+
},
|
| 9 |
+
"action_out_proj_cfg": {
|
| 10 |
+
"_target_": "torch.nn.Linear"
|
| 11 |
+
},
|
| 12 |
+
"action_space_cfg": {
|
| 13 |
+
"_target_": "alpamayo_r1.action_space.UnicycleAccelCurvatureActionSpace",
|
| 14 |
+
"a_lambda": 0.0001,
|
| 15 |
+
"a_ridge": 0.0001,
|
| 16 |
+
"accel_bounds": [
|
| 17 |
+
-9.8,
|
| 18 |
+
9.8
|
| 19 |
+
],
|
| 20 |
+
"accel_mean": 0.02902694707164455,
|
| 21 |
+
"accel_std": 0.6810426736454882,
|
| 22 |
+
"curvature_bounds": [
|
| 23 |
+
-0.33,
|
| 24 |
+
0.33
|
| 25 |
+
],
|
| 26 |
+
"curvature_mean": 0.0002692167976330542,
|
| 27 |
+
"curvature_std": 0.026148280660833106,
|
| 28 |
+
"dt": 0.1,
|
| 29 |
+
"kappa_lambda": 0.0001,
|
| 30 |
+
"kappa_ridge": 0.0001,
|
| 31 |
+
"n_waypoints": 64,
|
| 32 |
+
"theta_lambda": 1e-06,
|
| 33 |
+
"theta_ridge": 1e-08,
|
| 34 |
+
"v_lambda": 1e-06,
|
| 35 |
+
"v_ridge": 0.0001
|
| 36 |
+
},
|
| 37 |
+
"add_special_tokens": true,
|
| 38 |
+
"architectures": [
|
| 39 |
+
"ExpertModel"
|
| 40 |
+
],
|
| 41 |
+
"attn_implementation": "flash_attention_2",
|
| 42 |
+
"cotrain_vlm": false,
|
| 43 |
+
"diffusion_cfg": {
|
| 44 |
+
"_target_": "alpamayo_r1.diffusion.flow_matching.FlowMatching",
|
| 45 |
+
"inference_guidance_weight": 3.0,
|
| 46 |
+
"int_method": "euler",
|
| 47 |
+
"train_ignore_guidance_rate": 0.1,
|
| 48 |
+
"train_timestep_sampler": "beta",
|
| 49 |
+
"use_classifier_free_guidance": false,
|
| 50 |
+
"x_dims": "???"
|
| 51 |
+
},
|
| 52 |
+
"dtype": "bfloat16",
|
| 53 |
+
"expert_cfg": {
|
| 54 |
+
"dtype": "bfloat16",
|
| 55 |
+
"head_dim": 128,
|
| 56 |
+
"hidden_size": 2048,
|
| 57 |
+
"intermediate_size": 8256,
|
| 58 |
+
"num_attention_heads": 16
|
| 59 |
+
},
|
| 60 |
+
"expert_hist_traj_tokenizer_cfg": null,
|
| 61 |
+
"expert_non_causal_attention": true,
|
| 62 |
+
"hist_traj_embed_cfg": null,
|
| 63 |
+
"hist_traj_tokenizer_cfg": {
|
| 64 |
+
"_target_": "alpamayo_r1.models.delta_tokenizer.DeltaTrajectoryTokenizer"
|
| 65 |
+
},
|
| 66 |
+
"image_height": 320,
|
| 67 |
+
"image_width": 512,
|
| 68 |
+
"include_camera_ids": false,
|
| 69 |
+
"keep_same_dtype": true,
|
| 70 |
+
"legacy_inference_image_input_format": false,
|
| 71 |
+
"loss_weights": {
|
| 72 |
+
"future_traj": 1.0,
|
| 73 |
+
"others": 1.0
|
| 74 |
+
},
|
| 75 |
+
"max_pixels": 196608,
|
| 76 |
+
"min_pixels": 163840,
|
| 77 |
+
"model_dtype": "bfloat16",
|
| 78 |
+
"model_type": "alpamayo_r1",
|
| 79 |
+
"stop_grad_from_vlm": true,
|
| 80 |
+
"tokens_per_future_traj": 128,
|
| 81 |
+
"tokens_per_history_traj": 48,
|
| 82 |
+
"traj_loss_weight": 1.0,
|
| 83 |
+
"traj_token_ids": {
|
| 84 |
+
"future": 155685,
|
| 85 |
+
"future_end": 155683,
|
| 86 |
+
"future_start": 155681,
|
| 87 |
+
"history": 155684,
|
| 88 |
+
"history_end": 155676,
|
| 89 |
+
"history_start": 155674
|
| 90 |
+
},
|
| 91 |
+
"traj_token_start_idx": 151669,
|
| 92 |
+
"traj_tokenizer_cfg": {
|
| 93 |
+
"_recursive_": false,
|
| 94 |
+
"_target_": "alpamayo_r1.action_space.discrete_action_space.DiscreteTrajectoryTokenizer",
|
| 95 |
+
"action_space_cfg": {
|
| 96 |
+
"_target_": "alpamayo_r1.action_space.UnicycleAccelCurvatureActionSpace",
|
| 97 |
+
"a_lambda": 0.0001,
|
| 98 |
+
"a_ridge": 0.0001,
|
| 99 |
+
"accel_bounds": [
|
| 100 |
+
-9.8,
|
| 101 |
+
9.8
|
| 102 |
+
],
|
| 103 |
+
"accel_mean": 0.02902694707164455,
|
| 104 |
+
"accel_std": 0.6810426736454882,
|
| 105 |
+
"curvature_bounds": [
|
| 106 |
+
-0.33,
|
| 107 |
+
0.33
|
| 108 |
+
],
|
| 109 |
+
"curvature_mean": 0.0002692167976330542,
|
| 110 |
+
"curvature_std": 0.026148280660833106,
|
| 111 |
+
"dt": 0.1,
|
| 112 |
+
"kappa_lambda": 0.0001,
|
| 113 |
+
"kappa_ridge": 0.0001,
|
| 114 |
+
"n_waypoints": 64,
|
| 115 |
+
"theta_lambda": 1e-06,
|
| 116 |
+
"theta_ridge": 1e-08,
|
| 117 |
+
"v_lambda": 1e-06,
|
| 118 |
+
"v_ridge": 0.0001
|
| 119 |
+
},
|
| 120 |
+
"dims_max": [
|
| 121 |
+
10,
|
| 122 |
+
10
|
| 123 |
+
],
|
| 124 |
+
"dims_min": [
|
| 125 |
+
-10,
|
| 126 |
+
-10
|
| 127 |
+
],
|
| 128 |
+
"num_bins": 3000
|
| 129 |
+
},
|
| 130 |
+
"traj_vocab_size": 4000,
|
| 131 |
+
"transformers_version": "4.57.1",
|
| 132 |
+
"vlm_backend": "qwenvl3",
|
| 133 |
+
"vocab_size": 155697
|
| 134 |
+
}
|
model-00001-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6aabd8d143cff0295a60b515dfcb5ba6a5b1b5acf7cea1d6c6254ed653d35965
|
| 3 |
+
size 4928204944
|
model-00002-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e346fcf2bf4ebd75853bd17b5744c4158cf621aba195aad7c7e2f1e484c1ae20
|
| 3 |
+
size 4915963032
|
model-00003-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7acad9950402f002825aa41048155ea1a8a2fb5f7a00501154d52828291171c9
|
| 3 |
+
size 4983071160
|
model-00004-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:59f3ec7a6983ae2654b7ffe5d79f59d3d523b8a71a0ffc83362816886636ac0a
|
| 3 |
+
size 4980341192
|
model-00005-of-00005.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5753b1fd57638b70db4882cc8061eedfa28c2748a37414eb4032e20b1bb2e4c1
|
| 3 |
+
size 2349614880
|
model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|