Upload 12 files

0.codes.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23343f45a9150f5c1654f8919d1487c82ba5eb5971c269f88776ade07d356351
+size 1753500

0.metadata.json

@@ -0,0 +1,6 @@
+{
+    "passage_offset": 0,
+    "num_passages": 2716,
+    "num_embeddings": 438080,
+    "embedding_offset": 0
+}

0.residuals.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b74beb17fe3495b290af8aad966bf1ac9311208d220829d9fe7874ba2c843785
+size 56075440

avg_residual.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c0a2275675caa1b5ed953f08b96bd4e721ed69fc52564c43c908c454f9afc60
+size 1205

buckets.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0241a8c7f52deff2d87565c630997bde65d21b0480ffd869f66385a029b04b4c
+size 2904

centroids.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c87a0201a7b992741cc28c4a89bb46a94a0995a1105bc34814e46d783d3724d
+size 2098342

doclens.0.json

@@ -0,0 +1 @@
+[180,228,84,188,75,174,128,199,149,140,215,163,207,141,217,232,57,152,206,85,187,144,221,66,185,63,199,66,186,210,74,191,220,211,180,201,80,199,194,195,100,197,86,198,97,184,165,219,59,181,184,129,206,152,202,56,192,74,170,200,179,211,193,196,88,203,126,201,131,189,112,223,122,199,84,186,175,116,218,217,196,207,86,181,211,71,185,107,173,118,166,68,198,190,210,156,199,103,200,61,210,212,53,179,200,216,184,100,188,106,194,217,62,186,180,145,181,198,164,200,205,101,196,186,189,220,65,172,202,117,207,72,166,163,209,157,189,217,53,214,146,233,65,197,100,213,91,156,208,163,207,147,219,210,174,195,63,213,116,179,95,199,178,153,200,209,81,220,71,226,215,218,201,112,214,102,173,164,186,100,206,133,183,131,152,178,95,197,69,155,177,195,145,176,205,167,98,203,80,180,188,212,106,196,216,193,102,209,207,65,226,168,78,213,133,204,213,195,126,217,57,213,154,139,214,57,183,207,90,209,175,200,91,202,176,196,66,208,106,213,200,123,191,191,125,209,74,192,97,205,56,206,196,80,170,185,66,180,120,201,159,219,78,223,61,183,193,189,180,212,121,183,160,213,115,173,98,207,50,221,233,205,92,215,183,69,189,111,171,178,210,177,195,150,191,88,211,133,202,214,66,214,80,220,171,192,185,79,227,168,193,143,138,216,197,217,190,65,111,158,213,90,189,193,190,85,211,91,186,98,206,151,208,208,192,84,210,193,132,201,214,86,195,110,176,119,203,59,204,81,172,147,178,204,174,198,93,204,139,222,230,160,225,137,169,108,218,201,219,105,164,188,215,85,209,59,220,205,123,172,168,182,207,180,105,193,201,194,191,139,139,164,220,111,195,178,92,217,189,215,183,218,62,178,121,207,63,209,204,216,77,182,176,98,182,80,172,196,119,199,131,202,221,210,195,192,84,214,208,110,199,81,173,173,224,104,200,104,215,183,189,120,154,166,91,198,210,121,169,219,188,199,212,182,80,203,197,85,218,64,193,92,214,213,125,199,202,51,218,185,133,200,78,169,217,202,199,177,118,157,134,192,129,200,87,192,197,79,163,203,212,189,68,176,109,191,87,203,201,75,198,195,106,200,136,190,199,201,200,162,117,183,213,78,218,53,203,122,205,107,175,141,221,206,93,205,78,176,113,189,85,173,137,209,52,171,179,102,211,101,193,206,112,219,98,208,82,195,92,195,198,124,199,71,162,106,211,211,184,90,212,169,115,193,30,140,207,199,112,208,59,195,145,182,209,193,48,188,213,62,222,184,71,212,123,217,215,92,207,222,212,222,70,201,177,186,173,205,69,222,182,187,134,190,224,194,113,210,69,219,195,201,199,103,183,72,199,214,210,210,217,100,198,224,109,179,205,54,204,192,113,196,97,202,99,197,83,168,216,86,201,151,186,101,188,169,159,156,201,201,57,146,186,220,85,205,92,210,234,198,208,63,186,176,210,63,215,137,180,139,191,223,106,160,157,181,118,205,184,121,206,95,212,78,203,65,212,129,206,208,52,200,122,195,142,216,54,196,130,211,202,66,216,171,160,131,133,167,201,190,178,114,200,74,179,78,210,195,101,199,208,105,188,91,189,88,165,217,110,205,142,210,52,165,132,202,61,206,57,209,211,53,183,111,174,161,183,131,139,197,106,207,53,197,69,226,33,153,185,207,209,63,196,100,217,81,199,175,135,202,79,187,218,204,226,191,203,95,155,213,75,202,110,213,82,214,200,117,178,99,208,91,124,133,212,59,177,146,191,83,186,103,201,155,212,156,197,146,171,189,93,179,181,187,106,226,43,208,76,164,194,102,187,154,206,90,225,218,217,186,96,200,124,200,193,137,176,143,184,209,57,228,78,168,91,206,218,223,87,173,201,129,205,75,183,176,196,72,190,78,215,193,167,172,214,177,188,194,78,216,40,189,99,181,132,207,217,121,196,136,212,185,128,219,92,205,199,196,59,189,89,194,101,206,188,146,164,208,208,185,99,135,178,97,198,127,199,175,219,101,202,221,44,201,42,202,61,159,205,131,176,105,178,96,231,110,190,80,234,195,220,101,209,203,91,195,86,202,91,208,196,96,210,69,211,176,206,191,105,205,198,97,205,114,195,72,196,78,218,202,103,201,85,187,199,125,215,216,201,127,220,143,192,77,179,99,196,115,202,89,198,218,128,195,191,204,82,166,132,201,75,187,197,73,208,43,211,79,213,114,208,201,87,196,114,206,175,214,102,216,71,205,205,138,202,96,220,205,71,214,97,187,124,195,110,200,146,190,88,213,127,208,83,167,199,156,169,204,94,194,124,217,93,180,210,206,69,189,202,213,101,212,172,114,218,217,204,190,73,204,201,201,111,215,73,220,161,201,218,218,99,220,202,208,62,182,66,208,208,195,162,219,142,142,148,189,189,101,211,85,209,211,77,209,104,205,70,205,147,214,146,182,105,214,69,204,124,190,222,177,125,190,186,151,163,93,195,130,200,210,207,220,73,201,67,231,205,81,195,203,185,205,170,196,209,123,180,209,101,216,199,137,187,55,193,85,176,201,81,190,167,202,85,202,82,178,150,207,217,164,115,203,87,196,102,174,200,70,197,103,152,156,204,111,214,95,219,51,191,84,221,219,197,201,182,195,74,218,69,210,200,202,94,211,133,183,103,145,190,197,194,140,206,87,206,214,85,190,79,221,70,202,94,206,197,111,184,215,87,185,119,188,219,74,146,185,194,81,163,187,210,204,59,175,222,79,209,75,199,219,202,136,198,77,192,125,197,72,160,180,81,202,218,170,151,151,164,190,190,84,215,68,174,167,215,77,183,116,207,131,196,47,198,218,99,164,182,117,193,103,212,82,203,68,192,111,210,153,206,180,209,63,225,215,122,170,98,189,102,191,206,111,204,67,208,211,221,77,190,213,73,188,184,214,182,124,192,102,147,219,71,197,228,84,129,222,196,97,219,103,204,177,109,206,162,179,113,208,196,70,206,168,192,85,196,80,181,204,71,227,203,88,224,226,100,210,45,191,118,200,154,199,143,190,85,214,70,207,96,209,99,183,182,208,77,213,199,223,147,203,202,207,79,192,74,213,121,201,94,209,78,158,193,78,147,210,204,93,205,62,186,200,200,197,42,216,147,203,87,223,162,216,94,219,87,169,211,93,182,212,86,204,122,176,105,206,94,166,222,185,126,172,223,192,232,158,182,209,136,224,123,192,48,175,69,212,155,105,215,89,201,77,219,101,192,186,135,162,204,152,207,60,185,127,182,67,191,170,203,176,190,109,211,173,114,183,194,80,189,140,206,208,181,118,208,56,212,211,78,128,203,95,189,124,187,140,210,77,200,96,185,200,115,156,150,209,196,41,181,135,203,102,203,203,133,181,200,109,216,73,205,142,197,73,206,114,200,144,190,183,187,90,143,202,200,75,195,196,87,201,184,85,201,92,211,131,175,72,195,212,217,199,76,187,185,108,201,85,141,118,188,212,123,201,146,202,77,178,202,79,187,97,174,112,218,101,204,180,149,172,106,178,142,218,49,219,185,132,186,208,70,195,125,214,84,211,179,196,132,200,99,199,189,103,190,115,225,131,178,189,192,103,150,215,196,197,111,211,47,210,102,197,73,201,208,100,224,60,214,208,53,191,72,200,171,201,86,183,183,208,120,178,84,203,165,203,190,138,222,108,179,213,149,170,202,189,103,188,110,165,191,220,180,188,61,229,199,122,134,188,69,197,92,201,76,207,71,194,73,197,69,198,192,184,216,138,203,201,155,190,101,187,87,185,100,190,211,191,87,208,169,203,164,163,116,157,209,90,188,115,196,34,206,64,178,175,178,161,126,198,182,226,56,212,109,213,200,188,91,190,97,214,200,124,225,200,227,193,64,224,110,189,95,215,195,145,204,75,189,196,172,206,107,148,185,90,191,180,70,202,65,205,83,223,179,167,178,178,225,183,73,207,67,211,48,171,211,96,164,208,68,163,203,218,206,117,216,177,187,153,143,183,96,221,173,197,99,199,199,127,222,195,105,213,201,112,187,92,194,168,212,92,198,230,196,98,174,201,70,200,97,226,62,207,130,201,200,95,219,140,209,57,160,179,199,101,200,190,77,163,199,156,187,153,212,209,83,195,76,213,195,127,206,206,57,207,83,215,206,112,118,218,119,206,103,179,81,212,214,88,212,106,183,119,203,156,182,200,95,212,187,175,205,111,217,114,177,199,160,207,85,211,74,201,131,214,41,202,194,209,167,188,198,195,76,210,205,207,199,105,177,109,221,195,105,222,203,116,138,216,82,188,104,209,95,190,86,141,203,169,189,171,214,182,183,207,58,143,193,221,86,218,212,203,144,176,129,157,168,204,119,198,69,180,102,192,116,201,202,229,71,207,104,168,159,179,128,201,104,200,68,176,86,213,198,203,73,194,129,212,182,211,221,195,102,192,199,85,225,63,204,190,154,200,131,177,187,119,186,79,181,104,178,212,77,195,77,187,211,194,181,112,166,121,219,204,91,211,126,205,160,193,105,133,179,179,103,218,104,228,113,208,20,172,207,60,212,83,189,198,196,222,128,213,88,165,213,91,207,209,153,163,165,128,216,87,178,160,223,171,199,101,190,121,190,186,120,162,187,97,231,219,77,182,197,130,206,94,189,57,207,199,214,61,186,120,161,162,132,228,90,186,180,186,192,91,222,58,218,198,182,212,190,199,69,198,113,211,90,210,216,95,204,217,165,215,96,198,101,191,199,184,185,102,209,116,198,97,211,104,143,212,108,199,88,192,202,216,214,211,222,78,174,53,213,223,215,173,218,216,109,201,143,198,154,222,95,225,106,188,207,210,165,199,216,83,189,148,219,65,217,108,213,204,74,205,205,70,216,180,124,180,66,202,57,193,184,91,181,161,209,135,213,140,176,206,63,174,127,145,220,107,205,106,181,144,173,217,210,72,221,210,163,211,82,207,140,205,92,187,210,57,174,145,197,33,185,87,211,60,171,137,192,202,70,186,138,210,118,228,180,197,189,102,179,103,210,87,183,219,206,74,195,204,67,176,145,209,149,183,194,177,221,79,217,216,209,205,53,219,199,98,208,208,56,176,195,120,186,128,193,80,196,117,173,96,204,72,212,221,42,179,212,194,99,184,201,211,171,207,99,204,122,194,54,198,121,199,203,169,197,90,222,60,190,113,181,82,186,203,194,42,207,65,194,122,174,99,216,106,190,190,146,147,188,174,118,219,197,75,176,199,120,189,187,88,172,111,221,196,181,104,201,85,186,46,210,97,208,101,212,93,204,100,172,142,217,121,201,84,215,134,203,58,160,86,214,102,177,184,91,217,60,195,101,178,90,204,78,108,177,202,169,196,105,196,92,196,186,105,181,97,152,162,219,108,186,160,183,99,189,75,192,51,189,88,190,96,186,176,194,66,205,49,201,221,100,217,54,172,198,153,184,184,209,197,211,221,212,201,95,209,181,203,114,220,108,203,36,216,84,207,215,216,141,223,190,109,198,79,214,126,206,126,204,112,178,182,175,203,192,228,207,24,229,153,208,186,75,179,107,197,205,162,193,185,210,184,88,208,101,172,65,154,180,61,213,176,109,208,172,122,183,163,187,207,129,195,92,221,90,207,104,196,114,214,186,112,188,222,198,193,93,232,209,216,104,205,177,87,218,188,85,210,166,206,136,172,111,196,198,197,76,161,188,64,193,62,231,206,109,112,202,104,200,123,199,193,82,184,86,223,92,205,183,224,196,90,198,65,195,84,182,85,212,152,90,199,79,208,99,195,193,202,158,196,104,220,171,102,191,194,200,219,183,75,196,214,200,175,69,198,160,132,194,129,165,93,199,61,208,211,126,220,63,194,127,205,78,213,68,169,86,219,207,115,169,128,218,176,88,193,118,174,121,173,141,172,200,189,100,222,126,198,205,81,157,224,190,83,202,73,193,158,185,202,93,142,187,126,195,103,187,133,172,184,102,208,105,185,129,193,65,215,211,56,219,62,184,187,95,190,94,183,92,209,72,187,156,191,192,217,137,223,202,201,227,148,191,80,184,79,197,122,195,80,191,140,212,220,68,208,113,139,186,136]

ivf.pid.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:069fb40e49cdf2c93a2a981f91996014262b251ffe037e9c8840520ea89d80cf
+size 1166104

metadata.json

@@ -0,0 +1,70 @@
+{
+  "config":{
+    "query_token_id":"[unused0]",
+    "doc_token_id":"[unused1]",
+    "query_token":"[Q]",
+    "doc_token":"[D]",
+    "ncells":null,
+    "centroid_score_threshold":null,
+    "ndocs":null,
+    "load_index_with_mmap":false,
+    "index_path":null,
+    "index_bsize":32,
+    "nbits":8,
+    "kmeans_niters":20,
+    "resume":false,
+    "similarity":"cosine",
+    "bsize":64,
+    "accumsteps":1,
+    "lr":0.00001,
+    "maxsteps":400000,
+    "save_every":null,
+    "warmup":20000,
+    "warmup_bert":null,
+    "relu":false,
+    "nway":64,
+    "use_ib_negatives":true,
+    "reranker":false,
+    "distillation_alpha":1.0,
+    "ignore_scores":false,
+    "model_name":null,
+    "query_maxlen":32,
+    "attend_to_mask_tokens":false,
+    "interaction":"colbert",
+    "dim":128,
+    "doc_maxlen":256,
+    "mask_punctuation":true,
+    "checkpoint":"colbert-ir/colbertv2.0",
+    "triples":"/future/u/okhattab/root/unit/experiments/2021.10/downstream.distillation.round2.2_score/round2.nway6.cosine.ib/examples.64.json",
+    "collection":[
+      "list with 2716 elements starting with...",
+      [
+        "The reflective nature of the human eye is an underappreciated source of\ninformation about what the world around us looks like. By imaging the eyes of a\nmoving person, we can collect multiple views of a scene outside the camera's\ndirect line of sight through the reflections in the eyes. In this paper, we\nreconstruct a 3D scene beyond the camera's line of sight using portrait images\ncontaining eye reflections. This task is challenging due to 1) the difficulty\nof accurately estimating eye poses and 2) the entangled appearance of the eye\niris and the scene reflections. Our method jointly refines the cornea poses,\nthe radiance field depicting the scene, and the observer's eye iris texture. We\nfurther propose a simple regularization prior on the iris texture pattern to\nimprove reconstruction quality. Through various experiments on synthetic and\nreal-world captures featuring people with varied eye colors, we demonstrate the\nfeasibility of our approach to recover 3D scenes using eye reflections.",
+        "The recovery of occluded human meshes presents challenges for current methods\ndue to the difficulty in extracting effective image features under severe\nocclusion. In this paper, we introduce DPMesh, an innovative framework for\noccluded human mesh recovery that capitalizes on the profound diffusion prior\nabout object structure and spatial relationships embedded in a pre-trained\ntext-to-image diffusion model. Unlike previous methods reliant on conventional\nbackbones for vanilla feature extraction, DPMesh seamlessly integrates the\npre-trained denoising U-Net with potent knowledge as its image backbone and\nperforms a single-step inference to provide occlusion-aware information. To\nenhance the perception capability for occluded poses, DPMesh incorporates\nwell-designed guidance via condition injection, which produces effective\ncontrols from 2D observations for the denoising U-Net. Furthermore, we explore\na dedicated noisy key-point reasoning approach to mitigate disturbances arising\nfrom occlusion and crowded scenarios. This strategy fully unleashes the\nperceptual capability of the diffusion prior, thereby enhancing accuracy.\nExtensive experiments affirm the efficacy of our framework, as we outperform\nstate-of-the-art methods on both occlusion-specific and standard datasets.",
+        "This strategy fully unleashes the\nperceptual capability of the diffusion prior, thereby enhancing accuracy.\nExtensive experiments affirm the efficacy of our framework, as we outperform\nstate-of-the-art methods on both occlusion-specific and standard datasets. The\npersuasive results underscore its ability to achieve precise and robust 3D\nhuman mesh recovery, particularly in challenging scenarios involving occlusion\nand crowded scenes."
+      ]
+    ],
+    "queries":"/future/u/okhattab/data/MSMARCO/queries.train.tsv",
+    "index_name":"/content/my_index",
+    "overwrite":false,
+    "root":".ragatouille/",
+    "experiment":"colbert",
+    "index_root":null,
+    "name":"2024-06/16/21.05.35",
+    "rank":0,
+    "nranks":1,
+    "amp":true,
+    "gpus":1,
+    "avoid_fork_if_possible":false
+  },
+  "num_chunks":1,
+  "num_partitions":8192,
+  "num_embeddings":438080,
+  "avg_doclen":161.2960235641,
+  "RAGatouille":{
+    "index_config":{
+      "index_type":"PLAID",
+      "index_name":"/content/my_index"
+    }
+  }
+}

plan.json

@@ -0,0 +1,64 @@
+{
+    "config": {
+        "query_token_id": "[unused0]",
+        "doc_token_id": "[unused1]",
+        "query_token": "[Q]",
+        "doc_token": "[D]",
+        "ncells": null,
+        "centroid_score_threshold": null,
+        "ndocs": null,
+        "load_index_with_mmap": false,
+        "index_path": null,
+        "index_bsize": 32,
+        "nbits": 8,
+        "kmeans_niters": 20,
+        "resume": false,
+        "similarity": "cosine",
+        "bsize": 64,
+        "accumsteps": 1,
+        "lr": 1e-5,
+        "maxsteps": 400000,
+        "save_every": null,
+        "warmup": 20000,
+        "warmup_bert": null,
+        "relu": false,
+        "nway": 64,
+        "use_ib_negatives": true,
+        "reranker": false,
+        "distillation_alpha": 1.0,
+        "ignore_scores": false,
+        "model_name": null,
+        "query_maxlen": 32,
+        "attend_to_mask_tokens": false,
+        "interaction": "colbert",
+        "dim": 128,
+        "doc_maxlen": 256,
+        "mask_punctuation": true,
+        "checkpoint": "colbert-ir\/colbertv2.0",
+        "triples": "\/future\/u\/okhattab\/root\/unit\/experiments\/2021.10\/downstream.distillation.round2.2_score\/round2.nway6.cosine.ib\/examples.64.json",
+        "collection": [
+            "list with 2716 elements starting with...",
+            [
+                "The reflective nature of the human eye is an underappreciated source of\ninformation about what the world around us looks like. By imaging the eyes of a\nmoving person, we can collect multiple views of a scene outside the camera's\ndirect line of sight through the reflections in the eyes. In this paper, we\nreconstruct a 3D scene beyond the camera's line of sight using portrait images\ncontaining eye reflections. This task is challenging due to 1) the difficulty\nof accurately estimating eye poses and 2) the entangled appearance of the eye\niris and the scene reflections. Our method jointly refines the cornea poses,\nthe radiance field depicting the scene, and the observer's eye iris texture. We\nfurther propose a simple regularization prior on the iris texture pattern to\nimprove reconstruction quality. Through various experiments on synthetic and\nreal-world captures featuring people with varied eye colors, we demonstrate the\nfeasibility of our approach to recover 3D scenes using eye reflections.",
+                "The recovery of occluded human meshes presents challenges for current methods\ndue to the difficulty in extracting effective image features under severe\nocclusion. In this paper, we introduce DPMesh, an innovative framework for\noccluded human mesh recovery that capitalizes on the profound diffusion prior\nabout object structure and spatial relationships embedded in a pre-trained\ntext-to-image diffusion model. Unlike previous methods reliant on conventional\nbackbones for vanilla feature extraction, DPMesh seamlessly integrates the\npre-trained denoising U-Net with potent knowledge as its image backbone and\nperforms a single-step inference to provide occlusion-aware information. To\nenhance the perception capability for occluded poses, DPMesh incorporates\nwell-designed guidance via condition injection, which produces effective\ncontrols from 2D observations for the denoising U-Net. Furthermore, we explore\na dedicated noisy key-point reasoning approach to mitigate disturbances arising\nfrom occlusion and crowded scenarios. This strategy fully unleashes the\nperceptual capability of the diffusion prior, thereby enhancing accuracy.\nExtensive experiments affirm the efficacy of our framework, as we outperform\nstate-of-the-art methods on both occlusion-specific and standard datasets.",
+                "This strategy fully unleashes the\nperceptual capability of the diffusion prior, thereby enhancing accuracy.\nExtensive experiments affirm the efficacy of our framework, as we outperform\nstate-of-the-art methods on both occlusion-specific and standard datasets. The\npersuasive results underscore its ability to achieve precise and robust 3D\nhuman mesh recovery, particularly in challenging scenarios involving occlusion\nand crowded scenes."
+            ]
+        ],
+        "queries": "\/future\/u\/okhattab\/data\/MSMARCO\/queries.train.tsv",
+        "index_name": "\/content\/my_index",
+        "overwrite": false,
+        "root": ".ragatouille\/",
+        "experiment": "colbert",
+        "index_root": null,
+        "name": "2024-06\/16\/21.05.35",
+        "rank": 0,
+        "nranks": 1,
+        "amp": true,
+        "gpus": 1,
+        "avoid_fork_if_possible": false
+    },
+    "num_chunks": 1,
+    "num_partitions": 8192,
+    "num_embeddings_est": 438079.99169921875,
+    "avg_doclen_est": 161.2960205078125
+}