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

Add OpenAI text-embedding-3-large msmarco 2cr #1863

Merged
merged 3 commits into from
Apr 14, 2024
Merged

Add OpenAI text-embedding-3-large msmarco 2cr #1863

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
478124a
Add OpenAI text-embedding-3-large msmarco 2cr
manveertamber Apr 14, 2024
d692035
Increment number of MS MARCO faiss indexes in test
manveertamber Apr 14, 2024
a0a0f0e
Add index readme
manveertamber Apr 14, 2024
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
169 changes: 139 additions & 30 deletions docs/2cr/msmarco-v1-passage.html
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5941,11 +5941,120 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</div>
<!-- Tabs content -->

</div></td>
</tr>
<!-- Condition: OpenAI text-embedding-3-large: pre-encoded queries -->
<tr class="accordion-toggle collapsed" id="row52" data-toggle="collapse" data-parent="#row52" href="#collapse52">
<td class="expand-button"></td>
<td style="min-width: 85px"></td>
<td style="min-width: 400px">OpenAI text-embedding-3-large: pre-encoded queries</td>
<td>0.5259</td>
<td>0.7173</td>
<td>0.8991</td>
<td></td>
<td>0.5134</td>
<td>0.7163</td>
<td>0.8884</td>
<td></td>
<td>0.3342</td>
<td>0.9885</td>
</tr>
<tr class="hide-table-padding">
<td></td>
<td colspan="11">
<div id="collapse52" class="collapse in p-3">

<!-- Tabs navs -->
<ul class="nav nav-tabs mb-3" id="row52-tabs" role="tablist">
<li class="nav-item" role="presentation">
<a class="nav-link active" id="row52-tab1-header" data-mdb-toggle="tab" href="#row52-tab1" role="tab" aria-controls="row52-tab1" aria-selected="true" style="text-transform:none">TREC 2019</a>
</li>
<li class="nav-item" role="presentation">
<a class="nav-link" id="row52-tab2-header" data-mdb-toggle="tab" href="#row52-tab2" role="tab" aria-controls="row52-tab2" aria-selected="false" style="text-transform:none">TREC 2020</a>
</li>
<li class="nav-item" role="presentation">
<a class="nav-link" id="row52-tab3-header" data-mdb-toggle="tab" href="#row52-tab3" role="tab" aria-controls="row52-tab3" aria-selected="false" style="text-transform:none">dev</a>
</li>
</ul>
<!-- Tabs navs -->

<!-- Tabs content -->
<div class="tab-content" id="row52-content">
<div class="tab-pane fade show active" id="row52-tab1" role="tabpanel" aria-labelledby="row52-tab1">
Command to generate run on TREC 2019 queries:

<blockquote class="mycode">
<pre><code>python -m pyserini.search.faiss \
--threads 16 --batch-size 512 \
--index msmarco-v1-passage.openai-text-embedding-3-large \
--topics dl19-passage --encoded-queries openai-text-embedding-3-large-dl19-passage \
--output run.msmarco-v1-passage.openai-text-embedding-3-large.dl19.txt
</code></pre></blockquote>
Evaluation commands:

<blockquote class="mycode">
<pre><code>python -m pyserini.eval.trec_eval -c -l 2 -m map dl19-passage \
run.msmarco-v1-passage.openai-text-embedding-3-large.dl19.txt
python -m pyserini.eval.trec_eval -c -m ndcg_cut.10 dl19-passage \
run.msmarco-v1-passage.openai-text-embedding-3-large.dl19.txt
python -m pyserini.eval.trec_eval -c -l 2 -m recall.1000 dl19-passage \
run.msmarco-v1-passage.openai-text-embedding-3-large.dl19.txt
</code></pre>
</blockquote>

</div>
<div class="tab-pane fade" id="row52-tab2" role="tabpanel" aria-labelledby="row52-tab2">
Command to generate run on TREC 2020 queries:

<blockquote class="mycode">
<pre><code>python -m pyserini.search.faiss \
--threads 16 --batch-size 512 \
--index msmarco-v1-passage.openai-text-embedding-3-large \
--topics dl20 --encoded-queries openai-text-embedding-3-large-dl20 \
--output run.msmarco-v1-passage.openai-text-embedding-3-large.dl20.txt
</code></pre></blockquote>
Evaluation commands:

<blockquote class="mycode">
<pre><code>python -m pyserini.eval.trec_eval -c -l 2 -m map dl20-passage \
run.msmarco-v1-passage.openai-text-embedding-3-large.dl20.txt
python -m pyserini.eval.trec_eval -c -m ndcg_cut.10 dl20-passage \
run.msmarco-v1-passage.openai-text-embedding-3-large.dl20.txt
python -m pyserini.eval.trec_eval -c -l 2 -m recall.1000 dl20-passage \
run.msmarco-v1-passage.openai-text-embedding-3-large.dl20.txt
</code></pre>
</blockquote>

</div>
<div class="tab-pane fade" id="row52-tab3" role="tabpanel" aria-labelledby="row52-tab3">
Command to generate run on dev queries:

<blockquote class="mycode">
<pre><code>python -m pyserini.search.faiss \
--threads 16 --batch-size 512 \
--index msmarco-v1-passage.openai-text-embedding-3-large \
--topics msmarco-passage-dev-subset --encoded-queries openai-text-embedding-3-large-msmarco-passage-dev-subset \
--output run.msmarco-v1-passage.openai-text-embedding-3-large.dev.txt
</code></pre></blockquote>
Evaluation commands:

<blockquote class="mycode">
<pre><code>python -m pyserini.eval.trec_eval -c -M 10 -m recip_rank msmarco-passage-dev-subset \
run.msmarco-v1-passage.openai-text-embedding-3-large.dev.txt
python -m pyserini.eval.trec_eval -c -m recall.1000 msmarco-passage-dev-subset \
run.msmarco-v1-passage.openai-text-embedding-3-large.dev.txt
</code></pre>
</blockquote>

</div>
</div>
<!-- Tabs content -->

</div></td>
</tr>
<tr><td style="border-bottom: 0"></td></tr>
<!-- Condition: cosDPR-distil: PyTorch -->
<tr class="accordion-toggle collapsed" id="row52" data-toggle="collapse" data-parent="#row52" href="#collapse52">
<tr class="accordion-toggle collapsed" id="row53" data-toggle="collapse" data-parent="#row53" href="#collapse53">
<td class="expand-button"></td>
<td style="min-width: 85px">[13]</td>
<td style="min-width: 400px">cosDPR-distil: PyTorch</td>
Expand All @@ -5963,25 +6072,25 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
<tr class="hide-table-padding">
<td></td>
<td colspan="11">
<div id="collapse52" class="collapse in p-3">
<div id="collapse53" class="collapse in p-3">

<!-- Tabs navs -->
<ul class="nav nav-tabs mb-3" id="row52-tabs" role="tablist">
<ul class="nav nav-tabs mb-3" id="row53-tabs" role="tablist">
<li class="nav-item" role="presentation">
<a class="nav-link active" id="row52-tab1-header" data-mdb-toggle="tab" href="#row52-tab1" role="tab" aria-controls="row52-tab1" aria-selected="true" style="text-transform:none">TREC 2019</a>
<a class="nav-link active" id="row53-tab1-header" data-mdb-toggle="tab" href="#row53-tab1" role="tab" aria-controls="row53-tab1" aria-selected="true" style="text-transform:none">TREC 2019</a>
</li>
<li class="nav-item" role="presentation">
<a class="nav-link" id="row52-tab2-header" data-mdb-toggle="tab" href="#row52-tab2" role="tab" aria-controls="row52-tab2" aria-selected="false" style="text-transform:none">TREC 2020</a>
<a class="nav-link" id="row53-tab2-header" data-mdb-toggle="tab" href="#row53-tab2" role="tab" aria-controls="row53-tab2" aria-selected="false" style="text-transform:none">TREC 2020</a>
</li>
<li class="nav-item" role="presentation">
<a class="nav-link" id="row52-tab3-header" data-mdb-toggle="tab" href="#row52-tab3" role="tab" aria-controls="row52-tab3" aria-selected="false" style="text-transform:none">dev</a>
<a class="nav-link" id="row53-tab3-header" data-mdb-toggle="tab" href="#row53-tab3" role="tab" aria-controls="row53-tab3" aria-selected="false" style="text-transform:none">dev</a>
</li>
</ul>
<!-- Tabs navs -->

<!-- Tabs content -->
<div class="tab-content" id="row52-content">
<div class="tab-pane fade show active" id="row52-tab1" role="tabpanel" aria-labelledby="row52-tab1">
<div class="tab-content" id="row53-content">
<div class="tab-pane fade show active" id="row53-tab1" role="tabpanel" aria-labelledby="row53-tab1">
Command to generate run on TREC 2019 queries:

<blockquote class="mycode">
Expand All @@ -6005,7 +6114,7 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</blockquote>

</div>
<div class="tab-pane fade" id="row52-tab2" role="tabpanel" aria-labelledby="row52-tab2">
<div class="tab-pane fade" id="row53-tab2" role="tabpanel" aria-labelledby="row53-tab2">
Command to generate run on TREC 2020 queries:

<blockquote class="mycode">
Expand All @@ -6029,7 +6138,7 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</blockquote>

</div>
<div class="tab-pane fade" id="row52-tab3" role="tabpanel" aria-labelledby="row52-tab3">
<div class="tab-pane fade" id="row53-tab3" role="tabpanel" aria-labelledby="row53-tab3">
Command to generate run on dev queries:

<blockquote class="mycode">
Expand Down Expand Up @@ -6058,7 +6167,7 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</tr>
<tr><td style="border-bottom: 0"></td></tr>
<!-- Condition: BGE-base-en-v1.5: PyTorch -->
<tr class="accordion-toggle collapsed" id="row53" data-toggle="collapse" data-parent="#row53" href="#collapse53">
<tr class="accordion-toggle collapsed" id="row54" data-toggle="collapse" data-parent="#row54" href="#collapse54">
<td class="expand-button"></td>
<td style="min-width: 85px">[14]</td>
<td style="min-width: 400px">BGE-base-en-v1.5: PyTorch</td>
Expand All @@ -6076,25 +6185,25 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
<tr class="hide-table-padding">
<td></td>
<td colspan="11">
<div id="collapse53" class="collapse in p-3">
<div id="collapse54" class="collapse in p-3">

<!-- Tabs navs -->
<ul class="nav nav-tabs mb-3" id="row53-tabs" role="tablist">
<ul class="nav nav-tabs mb-3" id="row54-tabs" role="tablist">
<li class="nav-item" role="presentation">
<a class="nav-link active" id="row53-tab1-header" data-mdb-toggle="tab" href="#row53-tab1" role="tab" aria-controls="row53-tab1" aria-selected="true" style="text-transform:none">TREC 2019</a>
<a class="nav-link active" id="row54-tab1-header" data-mdb-toggle="tab" href="#row54-tab1" role="tab" aria-controls="row54-tab1" aria-selected="true" style="text-transform:none">TREC 2019</a>
</li>
<li class="nav-item" role="presentation">
<a class="nav-link" id="row53-tab2-header" data-mdb-toggle="tab" href="#row53-tab2" role="tab" aria-controls="row53-tab2" aria-selected="false" style="text-transform:none">TREC 2020</a>
<a class="nav-link" id="row54-tab2-header" data-mdb-toggle="tab" href="#row54-tab2" role="tab" aria-controls="row54-tab2" aria-selected="false" style="text-transform:none">TREC 2020</a>
</li>
<li class="nav-item" role="presentation">
<a class="nav-link" id="row53-tab3-header" data-mdb-toggle="tab" href="#row53-tab3" role="tab" aria-controls="row53-tab3" aria-selected="false" style="text-transform:none">dev</a>
<a class="nav-link" id="row54-tab3-header" data-mdb-toggle="tab" href="#row54-tab3" role="tab" aria-controls="row54-tab3" aria-selected="false" style="text-transform:none">dev</a>
</li>
</ul>
<!-- Tabs navs -->

<!-- Tabs content -->
<div class="tab-content" id="row53-content">
<div class="tab-pane fade show active" id="row53-tab1" role="tabpanel" aria-labelledby="row53-tab1">
<div class="tab-content" id="row54-content">
<div class="tab-pane fade show active" id="row54-tab1" role="tabpanel" aria-labelledby="row54-tab1">
Command to generate run on TREC 2019 queries:

<blockquote class="mycode">
Expand All @@ -6120,7 +6229,7 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</blockquote>

</div>
<div class="tab-pane fade" id="row53-tab2" role="tabpanel" aria-labelledby="row53-tab2">
<div class="tab-pane fade" id="row54-tab2" role="tabpanel" aria-labelledby="row54-tab2">
Command to generate run on TREC 2020 queries:

<blockquote class="mycode">
Expand All @@ -6146,7 +6255,7 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</blockquote>

</div>
<div class="tab-pane fade" id="row53-tab3" role="tabpanel" aria-labelledby="row53-tab3">
<div class="tab-pane fade" id="row54-tab3" role="tabpanel" aria-labelledby="row54-tab3">
Command to generate run on dev queries:

<blockquote class="mycode">
Expand Down Expand Up @@ -6177,7 +6286,7 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</tr>
<tr><td style="border-bottom: 0"></td></tr>
<!-- Condition: Cohere Embed English v3.0: pre-encoded queries -->
<tr class="accordion-toggle collapsed" id="row54" data-toggle="collapse" data-parent="#row54" href="#collapse54">
<tr class="accordion-toggle collapsed" id="row55" data-toggle="collapse" data-parent="#row55" href="#collapse55">
<td class="expand-button"></td>
<td style="min-width: 85px"></td>
<td style="min-width: 400px">Cohere Embed English v3.0: pre-encoded queries</td>
Expand All @@ -6195,25 +6304,25 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
<tr class="hide-table-padding">
<td></td>
<td colspan="11">
<div id="collapse54" class="collapse in p-3">
<div id="collapse55" class="collapse in p-3">

<!-- Tabs navs -->
<ul class="nav nav-tabs mb-3" id="row54-tabs" role="tablist">
<ul class="nav nav-tabs mb-3" id="row55-tabs" role="tablist">
<li class="nav-item" role="presentation">
<a class="nav-link active" id="row54-tab1-header" data-mdb-toggle="tab" href="#row54-tab1" role="tab" aria-controls="row54-tab1" aria-selected="true" style="text-transform:none">TREC 2019</a>
<a class="nav-link active" id="row55-tab1-header" data-mdb-toggle="tab" href="#row55-tab1" role="tab" aria-controls="row55-tab1" aria-selected="true" style="text-transform:none">TREC 2019</a>
</li>
<li class="nav-item" role="presentation">
<a class="nav-link" id="row54-tab2-header" data-mdb-toggle="tab" href="#row54-tab2" role="tab" aria-controls="row54-tab2" aria-selected="false" style="text-transform:none">TREC 2020</a>
<a class="nav-link" id="row55-tab2-header" data-mdb-toggle="tab" href="#row55-tab2" role="tab" aria-controls="row55-tab2" aria-selected="false" style="text-transform:none">TREC 2020</a>
</li>
<li class="nav-item" role="presentation">
<a class="nav-link" id="row54-tab3-header" data-mdb-toggle="tab" href="#row54-tab3" role="tab" aria-controls="row54-tab3" aria-selected="false" style="text-transform:none">dev</a>
<a class="nav-link" id="row55-tab3-header" data-mdb-toggle="tab" href="#row55-tab3" role="tab" aria-controls="row55-tab3" aria-selected="false" style="text-transform:none">dev</a>
</li>
</ul>
<!-- Tabs navs -->

<!-- Tabs content -->
<div class="tab-content" id="row54-content">
<div class="tab-pane fade show active" id="row54-tab1" role="tabpanel" aria-labelledby="row54-tab1">
<div class="tab-content" id="row55-content">
<div class="tab-pane fade show active" id="row55-tab1" role="tabpanel" aria-labelledby="row55-tab1">
Command to generate run on TREC 2019 queries:

<blockquote class="mycode">
Expand All @@ -6236,7 +6345,7 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</blockquote>

</div>
<div class="tab-pane fade" id="row54-tab2" role="tabpanel" aria-labelledby="row54-tab2">
<div class="tab-pane fade" id="row55-tab2" role="tabpanel" aria-labelledby="row55-tab2">
Command to generate run on TREC 2020 queries:

<blockquote class="mycode">
Expand All @@ -6259,7 +6368,7 @@ <h1 class="mb-3">MS MARCO V1 Passage</h1>
</blockquote>

</div>
<div class="tab-pane fade" id="row54-tab3" role="tabpanel" aria-labelledby="row54-tab3">
<div class="tab-pane fade" id="row55-tab3" role="tabpanel" aria-labelledby="row55-tab3">
Command to generate run on dev queries:

<blockquote class="mycode">
Expand Down
4 changes: 4 additions & 0 deletions docs/prebuilt-indexes.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -939,6 +939,10 @@ Detailed configuration information for the pre-built indexes are stored in [`pys
<dt></dt><b><code>msmarco-v1-passage.cohere-embed-english-v3.0</code></b>
<dd>Faiss FlatIP index of the MS MARCO passage corpus encoded by Cohere Embed English v3.0
</dd>
<dt></dt><b><code>msmarco-v1-passage.openai-text-embedding-3-large</code></b>
[<a href="../pyserini/resources/index-metadata/faiss-flat.msmarco-v1-passage.openai-text-embedding-3-large.20240410.c13cd6.README.md">readme</a>]
<dd>Faiss FlatIP index of the MS MARCO passage corpus encoded by OpenAI text-embedding-3-large
</dd>
<dt></dt><b><code>msmarco-v1-doc.ance-maxp</code></b>
<dd>Faiss FlatIP index of the MS MARCO document corpus encoded by the ANCE MaxP encoder
</dd>
Expand Down
23 changes: 23 additions & 0 deletions pyserini/2cr/msmarco-v1-passage.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1201,6 +1201,29 @@ conditions:
- MAP: 0.4938
nDCG@10: 0.6666
R@1K: 0.8919
- name: openai-text-embedding-3-large
display: "OpenAI text-embedding-3-large: pre-encoded queries"
display-html: "OpenAI text-embedding-3-large: pre-encoded queries"
display-row: ""
command: python -m pyserini.search.faiss --threads ${dense_threads} --batch-size ${dense_batch_size} --index msmarco-v1-passage.openai-text-embedding-3-large --topics $topics --encoded-queries openai-text-embedding-3-large-$topics --output $output
topics:
- topic_key: msmarco-passage-dev-subset
eval_key: msmarco-passage-dev-subset
scores:
- MRR@10: 0.3342
R@1K: 0.9885
- topic_key: dl19-passage
eval_key: dl19-passage
scores:
- MAP: 0.5259
nDCG@10: 0.7173
R@1K: 0.8991
- topic_key: dl20
eval_key: dl20-passage
scores:
- MAP: 0.5134
nDCG@10: 0.7163
R@1K: 0.8884
- name: cohere-embed-english-v3.0
display: "Cohere Embed English v3.0: pre-encoded queries"
display-html: "Cohere Embed English v3.0: pre-encoded queries"
Expand Down
1 change: 1 addition & 0 deletions pyserini/2cr/msmarco.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -103,6 +103,7 @@
'',
'openai-ada2',
'openai-ada2-hyde',
'openai-text-embedding-3-large',
'',
'cosdpr-distil-pytorch',
'',
Expand Down
30 changes: 30 additions & 0 deletions pyserini/encoded_query_info.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -515,6 +515,36 @@
"total_queries": 6980,
"downloaded": False
},
"openai-text-embedding-3-large-dl19-passage": {
"description": "TREC DL19 passage queries encoded by OpenAI text-embedding-3-large.",
"urls": [
"https://github.com/castorini/pyserini-data/raw/main/encoded-queries/query-embedding-openai-text-embedding-3-large-dl19-passage-20240410-c13cd6.tar.gz",
],
"md5": "a2e4ad9dc3288d97b77577552df9ee2b",
"size (bytes)": 541753,
"total_queries": 43,
"downloaded": False
},
"openai-text-embedding-3-large-dl20": {
"description": "TREC DL20 passage queries encoded by OpenAI text-embedding-3-large.",
"urls": [
"https://github.com/castorini/pyserini-data/raw/main/encoded-queries/query-embedding-openai-text-embedding-3-large-dl20-passage-20240410-c13cd6.tar.gz",
],
"md5": "9911fb1012ff5651f0cf832a81943967",
"size (bytes)": 2515768,
"total_queries": 200,
"downloaded": False
},
"openai-text-embedding-3-large-msmarco-passage-dev-subset": {
"description": "MS MARCO passage dev set queries encoded by OpenAI text-embedding-3-large.",
"urls": [
"https://github.com/castorini/pyserini-data/raw/main/encoded-queries/query-embedding-openai-text-embedding-3-large-msmarco-passage-dev-subset-20240410-c13cd6.tar.gz",
],
"md5": "4b0bce9c7cb0b55e49920d340924c92f",
"size (bytes)": 87687020,
"total_queries": 6980,
"downloaded": False
},
"atomic-v0.2.1-text-ViT-L-14.laion2b_s32b_b82k-validation": {
"description": "AToMiC text v0.2.1 validation set encoded by ViT-L-14.laion2b_s32b_b82k.",
"urls": [
Expand Down
13 changes: 13 additions & 0 deletions pyserini/prebuilt_index_info.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3344,6 +3344,19 @@
"downloaded": False,
"texts": "msmarco-v1-passage"
},
"msmarco-v1-passage.openai-text-embedding-3-large": {
"description": "Faiss FlatIP index of the MS MARCO passage corpus encoded by OpenAI text-embedding-3-large",
"filename": "faiss-flat.msmarco-v1-passage.openai-text-embedding-3-large.20240410.c13cd6.tar.gz",
"readme": "faiss-flat.msmarco-v1-passage.openai-text-embedding-3-large.20240410.c13cd6.README.md",
"urls": [
"https://rgw.cs.uwaterloo.ca/pyserini/indexes/faiss/faiss-flat.msmarco-v1-passage.openai-text-embedding-3-large.20240410.c13cd6.tar.gz"
],
"md5": "e52f046b1decc9bf3a55ac0ff70780d0",
"size compressed (bytes)": 87658796879,
"documents": 8841823,
"downloaded": False,
"texts": "msmarco-v1-passage"
},
"msmarco-v1-doc.ance-maxp": {
"description": "Faiss FlatIP index of the MS MARCO document corpus encoded by the ANCE MaxP encoder",
"filename": "faiss.msmarco-v1-doc.ance_maxp.20210304.b2a1b0.tar.gz",
Expand Down
Loading