README.md

metadata

base_model: sentence-transformers/all-MiniLM-L6-v2
language:
  - en
library_name: sentence-transformers
license: apache-2.0
metrics:
  - cosine_accuracy@1
  - cosine_accuracy@3
  - cosine_accuracy@5
  - cosine_accuracy@10
  - cosine_precision@1
  - cosine_precision@3
  - cosine_precision@5
  - cosine_precision@10
  - cosine_recall@1
  - cosine_recall@3
  - cosine_recall@5
  - cosine_recall@10
  - cosine_ndcg@10
  - cosine_mrr@10
  - cosine_map@100
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:4820
  - loss:MatryoshkaLoss
  - loss:MultipleNegativesRankingLoss
widget:
  - source_sentence: >-
      Defense 

      11 January 2024 ATP 3-21.8 5-57

       Reaction to enemy fires (for example, artillery and/or aviation) and
      CBRN

      attacks.

       Reports to higher, monitoring stockage levels, and cross leveling or
      resupply.

       CASEVAC and MEDEVAC procedures.

       Criteria to commitment the reserve.

      Figure 5-13. Main battle area (platoon engagements), example 

      FOLLOW THROUGH 

      5-167. During the planning for the defensive operation, the platoon leader
      must discern 

      from the company OPORD what the potential follow-on missions are and begin
      to plan

      how to achieve them. During this planning , the leader determines the
      possible timeline

      and location for defeat in detail , consolidate, reorganize, and
      transition which best

      f

      acilitates future operations and provides adequate protection.
    sentences:
      - What are some methods for distributing fires effectively in a platoon?
      - What should I consider when selecting battle positions for my unit?
      - What key factors are involved in planning for a defense scenario?
  - source_sentence: >-
      Chapter 6 

      6-24 ATP 3-21.8 11 January 2024 

      Figure 6-7. Movement to maneuver 

      TERRAIN 

      6-67. Platoon and squads enhance their own security during movement using
      covered

      and concealed terrain ; the use of the appropriate movement formation and
      movement

      technique; the actions taken to secure danger areas during crossing ; the
      enforcement of

      noise, light, and emissions control (for example, thermal and electronic)
      discipline; and

      us

      e of proper individual camouflage techniques. When planning and preparing
      for

      movement, leaders must consider how terrain affects security while
      simultaneously

      considering METT-TC (I). Some missions may require the platoon or
      individual squad

      to move on other than covered and concealed routes. While leaders may not
      be able t o

      pr

      event the unit ’s detection, they can ensure it moves on the battlefield
      in a time a nd

      pl

      ace for which the enemy is unprepared. Particularly when moving in the
      open, leaders

      must avoid predictability and continue to use terrain to their advantage.

      EXECUTION 

      6-68. During execution, leaders enforce camouflage discipline (Soldiers
      and their

      equipment). Leaders ensure the camouflage used by their Soldiers is
      appropriate to the

      terrain and season. Platoon standard operating procedures ( SOPs) specify
      elements of

      camouflage, noise and light discipline and emissions control; security
      halts; and actions

      at security halts.

      CAMOUFLAGE, NOISE, AND LIGHT DISCIPLINE AND EMISSIONS CONTROL 

      6-69. The platoon is visible to enemy forces and target acquisition
      capabilities on every

      spectrum, including visible light, sound, and across the electromagnetic
      spectrum. The
    sentences:
      - What is the process for preparing a machine gun range card?
      - How does terrain impact the security of a platoon during movement?
      - What’s the practical rate of fire for the M3 MAAWS?
  - source_sentence: >-
      Machine Gun Employment and Theory 

      11 January 2024 ATP 3-21.8 C-35

      SECURITY 

      C-115. Security includes all command measures to protect against surprise
      ,

      observation, and annoyance by the enemy. The gun team is responsible for
      its immediate 

      local security, specifically provided by the assistant gunner and/or
      ammunition bearer

      for close in local security to the gunner, who is fixated on deeper
      targets. Though the

      principal unit security measures against ground forces include employment
      of

      observation posts, security patrols, and detachments covering the front
      flanks and rear

      of the unit’s most valuable weapons systems and vulnerable areas. The
      composition and

      strength of these detachments depends on the size of the main body, its
      mission, and

      na

      ture of the opposition expected. The presence of machine guns with
      security

      detachments augments their firepower to delay, attack, and defend, by
      virtue of inherent

      firepower.

      C-116. The potential of air and ground attacks on the unit demands every
      possible

      precaution for maximum security while on the move. Where this situation
      exists , the

      machine gun crew must be thoroughly trained in the hasty delivery of
      antiaircraft fire

      and of counterfire against enemy ground forces. The distribution of the
      medium machine 

      guns in the formation is critical. The medium machine gun crew is
      constantly on the

      alert, particularly at halts , ready to deliver fire as soon as possible.
      If leaders expect a

      halt to exceed a brief period , they carefully choose medium machine gun
      positions to

      avoid unduly tiring the medium machine gun crew. If they expect the halt
      to extend for

      a long period, they can have the medium machine gun crew take up positions
      in support

      of the unit. The crew covers the direction from which they expect enemy
      activity as well

      as the direction from which the unit came. Leaders select positions
      permitting the

      delivery of fire in the most probable direction of enemy attack, such as
      valleys, draws,

      ridges, and spurs. They choose positions offering obstructed fire from
      potential enemy

      locations.

      EMPLOYMENT OF FIRE AND MOVEMENT 

      C-117. The employment of fire and movement is essential and greatly
      depends upon

      the other during maneuver. Without the support of covering fires ,
      maneuvering in the

      presence of enemy fire can result in disastrous losses. Covering fires ,
      especially

      providing fire superiority, allow maneuvering in the offense. However ,
      fire superiority

      alone rarely wins battles. The primary objective of the offense is to
      advance , occupy,

      and hold the enemy position.

      Machine Gun as a Base of Fire 

      C-118. Machine gun fire from a support by fire position must be the
      minimum possible

      to keep the enemy from returning fire. Ammunition must be conserved so the
      guns do

      not run out of ammunition.

      C-119. The weapons squad leader positions and controls the fires of all
      medium

      machine guns in the element. Machine gun targets include essential enemy
      weapons or

      groups of enemy targets either on the objective or attempting to reinforce
      or

      counterattack. In terms of engagement ranges, medium machine guns in the
      base-of-fire
    sentences:
      - How do observation posts aid in machine gun unit security?
      - >-
        Why would a unit choose to defend on a reverse slope rather than a
        forward slope?
      - >-
        What does the publication say about engagement area development for
        defense?
  - source_sentence: >-
      Chapter 7 

      7-18 ATP 3-21.8 11 January 2024 

      The PACE plan is a communication plan that exists for a specific mission
      or task, not a 

      specific unit, as the plan considers both intra- and inter-unit sharing of
      information. The 

      PACE plan designates the order in which an element will move through
      available 

      communications systems until contact can be established with the desired
      distant 

      element.  

      CHALLENGE AND PASSWORD OUTSIDE OF FRIENDLY LINES 

      7-61. The challenge and password from the signal operating instructions
      must not be

      used when the patrol is outside friendly lines. The unit ’s tactical SOP
      should state the

      procedure for establishing a patrol challenge and password as well as
      other combat

      identification features and patrol markings. Two methods for establishing
      a challenge

      and password are the odd number system and running password.

      Odd Number System 

      7-62. The leader specifies an odd number. The challenge can be any number
      less than

      the specified number. The password will be the number that must be added
      to it to equal

      the specified number, for example, the number is 9, the challenge is 4,
      and the password

      is 5.

      Running Password 

      7-63. Signal operating instructions also may designate a running password.
      This code

      word alerts a unit that friendly are approaching in a less than organized
      manner and

      possibly under pressure. The number of friendly approaching follows the
      running

      password. For example, if the running password is “eagle,” and seven
      friendl ies are

      a

      pproaching, they would say “eagle seven.”

      LOCATIONS OF KEY LEADERS 

      7-64. The patrol leader considers where best to locate throughout each
      phase of the

      patrol, and where to locate the APL , and other essential leaders for each
      phase of the

      patrol. The APL normally is with the following elements for each type of
      patrol:

       On a raid or ambush, the APL can be with the patrol leader on the
      objective

      or control the support element from the support position.

       On an area reconnaissance, the APL can move with one of the area

      reconnaissance elements or supervise security in the ORP.

       On a zone reconnaissance, the APL can move with one of the zone

      reconnaissance elements or move with the reconnaissance element setting up

      the linkup point.

      ACTIONS ON CHANCE CONTACT 

      7-65. The leader ’s plan must address actions on chance contact at each
      phase of the

      patrol. (See paragraphs 2-48 to 2-52 for additional information on actions
      on contact.)

      For the patrol’s mission the plan must address—
    sentences:
      - How does a platoon deal with obstacles during an assault?
      - >-
        What are some methods for setting up a challenge and password in the
        field?
      - >-
        What is the purpose of having one squad engage while others observe in
        an observed fire scenario?
  - source_sentence: >-
      Offense 

      11 January 2024 ATP 3-21.8 4-61

      light the target, making it easier to acquire effectively. Leaders and
      Soldiers 

      use the infrared devices to identify enemy or friendly personnel and then 

      engage targets using their aiming lights. 

      4-172. Illuminating rounds fired to burn on the ground can mark
      objectives. This helps

      the platoon orient on the objective but may adversely affect night vision
      devices.

      4-173. Leaders plan but do not always use illumination during limited
      visibility

      attacks. Battalion commanders normally control conventional illumination
      but ma y

      a

      uthorize the company commander to do so. If the commander decides to use

      conventional illumination , the commander should not call for it until the
      assault is

      initiated or the attack is detected. It should be placed on several
      locations over a wide

      area to confuse the enemy as to the exact place of the attack. It should
      be placed beyond

      the objective to help assaulting Soldiers see and fire at withdrawing or
      counterattacking

      enemy Soldiers. Infrared illumination is a good capability to light the
      objective without

      lighting it for enemy forces without night vision devices.  This advantage
      is degraded

      when used against a peer threat with the same night vision capabilities.

      4-174. The platoon leader , squad leaders , and vehicle commanders must
      know unit

      tactical SOP and develop sound COAs to synchronize the employment of
      infrared

      illumination devices , target designators , and aiming lights during their
      assault on the

      objective. These include using luminous tape or chemical lights to mark
      personnel and

      using weapons control restrictions.

      4-175. The platoon leader may use the following techniques to increase
      control during

      the assault:

       Use no flares, grenades, or obscuration on the objective.

       Use mortar or artillery rounds to orient attacking units.

       Use a base squad or fire team to pace and guide others.

       Reduce intervals between Soldiers and squads.

      4-176. Like a daylight attack , indirect and direct fires are planned for
      a limited

      visibility attack but are not executed unless the platoon is detected or
      is ready to assault.

      Some weapons may fire before the attack and maintain a pattern to deceive
      the enemy

      or to help cover noise ma de by the platoon ’s movement. This is not done
      if it will

      disclose the attack.

      4-177. Obscuration further reduces the enemy’s visibility, particularly if
      the enemy has

      night vision devices. The FO fires obscuration rounds close to or on enemy
      positions ,

      so it does not restrict friendly movement or hinder the reduction of
      obstacles. Employing 

      obscuration on the objective during the assault may make it hard for
      assaulting Soldiers

      to find enemy fighting positions. If enough thermal sights are available ,
      obscuration on

      the objective may provide a decisive advantage for a well-trained platoon.

      Note. I f the enemy is equipped with night vision devices , leaders must
      evaluate 

      the risk of using each technique and ensure the mission is not compromised
      by 

      the enemy’s ability to detect infrared light sources.
    sentences:
      - Can obscurants be used to hamper enemy fire support? How?
      - >-
        How can leaders effectively provide command and control during defensive
        operations?
      - What are the advantages of using infrared illumination in assaults?
model-index:
  - name: deep learning project 2
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 384
          type: dim_384
        metrics:
          - type: cosine_accuracy@1
            value: 0.0037313432835820895
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.013059701492537313
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.048507462686567165
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.4496268656716418
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.0037313432835820895
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.00435323383084577
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.009701492537313432
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.04496268656716418
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.0037313432835820895
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.013059701492537313
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.048507462686567165
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.4496268656716418
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.15012636108139818
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.06590188936271034
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.08623119999483674
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 256
          type: dim_256
        metrics:
          - type: cosine_accuracy@1
            value: 0.0037313432835820895
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.011194029850746268
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.03731343283582089
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.4458955223880597
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.0037313432835820895
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.003731343283582089
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.007462686567164179
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.04458955223880597
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.0037313432835820895
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.011194029850746268
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.03731343283582089
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.4458955223880597
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.14887734118005805
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.06525334636342103
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.08587360417470279
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 128
          type: dim_128
        metrics:
          - type: cosine_accuracy@1
            value: 0.0037313432835820895
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.009328358208955223
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.04664179104477612
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.43656716417910446
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.0037313432835820895
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.0031094527363184077
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.009328358208955225
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.043656716417910454
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.0037313432835820895
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.009328358208955223
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.04664179104477612
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.43656716417910446
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.14645163034094227
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.06459073679222935
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.08473376158047675
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 64
          type: dim_64
        metrics:
          - type: cosine_accuracy@1
            value: 0.0018656716417910447
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.007462686567164179
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.04291044776119403
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.4216417910447761
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.0018656716417910447
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.0024875621890547263
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.008582089552238806
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.04216417910447762
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.0018656716417910447
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.007462686567164179
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.04291044776119403
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.4216417910447761
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.13895211086835252
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.05946680289031035
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.07896404458930699
            name: Cosine Map@100

deep learning project 2

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L6-v2 on the json dataset. It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: sentence-transformers/all-MiniLM-L6-v2
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 384 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • json
• Language: en
• License: apache-2.0

Model Sources

• Documentation: Sentence Transformers Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
  (2): Normalize()
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("bbmb/deep-learning-for-embedding-model-ssilwal-qpham6_army_doc")
# Run inference
sentences = [
    'Offense \n11 January 2024 ATP 3-21.8 4-61\nlight the target, making it easier to acquire effectively. Leaders and Soldiers \nuse the infrared devices to identify enemy or friendly personnel and then \nengage targets using their aiming lights. \n4-172. Illuminating rounds fired to burn on the ground can mark objectives. This helps\nthe platoon orient on the objective but may adversely affect night vision devices.\n4-173. Leaders plan but do not always use illumination during limited visibility\nattacks. Battalion commanders normally control conventional illumination but ma y\na\nuthorize the company commander to do so. If the commander decides to use\nconventional illumination , the commander should not call for it until the assault is\ninitiated or the attack is detected. It should be placed on several locations over a wide\narea to confuse the enemy as to the exact place of the attack. It should be placed beyond\nthe objective to help assaulting Soldiers see and fire at withdrawing or counterattacking\nenemy Soldiers. Infrared illumination is a good capability to light the objective without\nlighting it for enemy forces without night vision devices.  This advantage is degraded\nwhen used against a peer threat with the same night vision capabilities.\n4-174. The platoon leader , squad leaders , and vehicle commanders must know unit\ntactical SOP and develop sound COAs to synchronize the employment of infrared\nillumination devices , target designators , and aiming lights during their assault on the\nobjective. These include using luminous tape or chemical lights to mark personnel and\nusing weapons control restrictions.\n4-175. The platoon leader may use the following techniques to increase control during\nthe assault:\n\uf06c Use no flares, grenades, or obscuration on the objective.\n\uf06c Use mortar or artillery rounds to orient attacking units.\n\uf06c Use a base squad or fire team to pace and guide others.\n\uf06c Reduce intervals between Soldiers and squads.\n4-176. Like a daylight attack , indirect and direct fires are planned for a limited\nvisibility attack but are not executed unless the platoon is detected or is ready to assault.\nSome weapons may fire before the attack and maintain a pattern to deceive the enemy\nor to help cover noise ma de by the platoon ’s movement. This is not done if it will\ndisclose the attack.\n4-177. Obscuration further reduces the enemy’s visibility, particularly if the enemy has\nnight vision devices. The FO fires obscuration rounds close to or on enemy positions ,\nso it does not restrict friendly movement or hinder the reduction of obstacles. Employing \nobscuration on the objective during the assault may make it hard for assaulting Soldiers\nto find enemy fighting positions. If enough thermal sights are available , obscuration on\nthe objective may provide a decisive advantage for a well-trained platoon.\nNote. I f the enemy is equipped with night vision devices , leaders must evaluate \nthe risk of using each technique and ensure the mission is not compromised by \nthe enemy’s ability to detect infrared light sources.',
    'What are the advantages of using infrared illumination in assaults?',
    'How can leaders effectively provide command and control during defensive operations?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Information Retrieval

• Datasets: dim_384, dim_256, dim_128 and dim_64
• Evaluated with InformationRetrievalEvaluator

Metric	dim_384	dim_256	dim_128	dim_64
cosine_accuracy@1	0.0037	0.0037	0.0037	0.0019
cosine_accuracy@3	0.0131	0.0112	0.0093	0.0075
cosine_accuracy@5	0.0485	0.0373	0.0466	0.0429
cosine_accuracy@10	0.4496	0.4459	0.4366	0.4216
cosine_precision@1	0.0037	0.0037	0.0037	0.0019
cosine_precision@3	0.0044	0.0037	0.0031	0.0025
cosine_precision@5	0.0097	0.0075	0.0093	0.0086
cosine_precision@10	0.045	0.0446	0.0437	0.0422
cosine_recall@1	0.0037	0.0037	0.0037	0.0019
cosine_recall@3	0.0131	0.0112	0.0093	0.0075
cosine_recall@5	0.0485	0.0373	0.0466	0.0429
cosine_recall@10	0.4496	0.4459	0.4366	0.4216
cosine_ndcg@10	0.1501	0.1489	0.1465	0.139
cosine_mrr@10	0.0659	0.0653	0.0646	0.0595
cosine_map@100	0.0862	0.0859	0.0847	0.079

Training Details

Training Dataset

json

• Dataset: json
• Size: 4,820 training samples
• Columns: positive and anchor
• Approximate statistics based on the first 1000 samples:

  	positive	anchor
  type	string	string
  details	min: 100 tokens mean: 248.18 tokens max: 256 tokens	min: 9 tokens mean: 15.06 tokens max: 27 tokens

• Samples:

  positive	anchor
  Appendix A A-22 ATP 3-21.8 11 January 2024 A-68. Observed fire. Usually is used when the platoon is in protected defensive positions with engagement ranges more than 2,500 meters for stabilized systems (when attached) and 1,500 meters for unstabilized systems. It can be employed between elements of the platoon, such as the squad lasing and observing while the weapons squad engages. The platoon leader directs one squad to engage. The remaining squads observe fires and prepare to engage on order in case the engaging element consistently misses its targets , experiences a malfunction, or runs low on ammunition. Observed fire allows for mutual observation and assistance while protecting the location of the observing elements. A-69. Sequential fire. Entails the subordinate elements of a unit engaging the same point or area target one after another in an arranged sequence. Sequential fire also can help to prevent the waste of ammunition, as when a platoon waits to see the effects of the ...	What is the purpose of having one squad engage while others observe in an observed fire scenario?
  Glossary Glossary-4 ATP 3-21.8 11 January 2024 PLD probable line of deployment PPEP personal protective equipment posture RFL restrictive fire line RM risk management ROE rules of engagement RS reduced sensitivity RTO radiotelephone operator S-2 battalion or brigade intelligence staff officer SALUTE size, activity, location, unit, time, and equipment SDM squad-designated marksman SITEMP situation template SLM shoulder-launched munition SOP standard operating procedure STP Soldier training publication TAA tactical assembly area TC training circular TCCC tactical combat casualty care TLP troop leading procedures TM technical manual TRP target reference point U.S. United States WARNORD warning order WCS weapons control status WP white phosphorous SECTION II – TERMS actions on contact A process to help leaders understand what is happening and to take action. (FM 3-90) air-ground operations The simultaneous or synchronized employment of ground forces with avi...	How is the term SDM used in the military?
  Chapter 1 1-2 ATP 3-21.8 11 January 2024 MISSION, CAPABILITIES, AND LIMITATIONS 1-2. The mission of the Infantry rifle platoon is to close with the enemy using fire and movement to destroy or capture enemy forces , or to repel enemy attacks by fire , close co mbat, and counterattack to control land areas , including populations and resources. The Infantry rifle platoon leader exercises command and control and directs the operation of the platoon and attached units while conducting combined arms warfare throughout the depth of the platoon’s area of operations (AO). Platoon missions , although not inclusive, may include reducing fortified areas , infiltrating and seizing objectives in the enemy’ s rear, eliminating enemy force remnants in restricted terrain , securing key facilities and activities, and conducting operations in support of stability operations tasks in the wake of maneuvering forces. Reconnaissance and surveillance operations and security operations remain a core compe...	What offensive and defensive actions can an Infantry rifle platoon perform?

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          384,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 16
• gradient_accumulation_steps: 8
• num_train_epochs: 20
• lr_scheduler_type: cosine
• warmup_ratio: 0.2
• bf16: True
• tf32: True
• load_best_model_at_end: True
• optim: adamw_torch_fused
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: epoch
• prediction_loss_only: True
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 16
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 8
• eval_accumulation_steps: None
• learning_rate: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 20
• max_steps: -1
• lr_scheduler_type: cosine
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.2
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• save_safetensors: True
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• no_cuda: False
• use_cpu: False
• use_mps_device: False
• seed: 42
• data_seed: None
• jit_mode_eval: False
• use_ipex: False
• bf16: True
• fp16: False
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: True
• local_rank: 0
• ddp_backend: None
• tpu_num_cores: None
• tpu_metrics_debug: False
• debug: []
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_prefetch_factor: None
• past_index: -1
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: True
• ignore_data_skip: False
• fsdp: []
• fsdp_min_num_params: 0
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• fsdp_transformer_layer_cls_to_wrap: None
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch_fused
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• ddp_find_unused_parameters: None
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• skip_memory_metrics: True
• use_legacy_prediction_loop: False
• push_to_hub: False
• resume_from_checkpoint: None
• hub_model_id: None
• hub_strategy: every_save
• hub_private_repo: False
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• eval_do_concat_batches: True
• fp16_backend: auto
• push_to_hub_model_id: None
• push_to_hub_organization: None
• mp_parameters:
• auto_find_batch_size: False
• full_determinism: False
• torchdynamo: None
• ray_scope: last
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• dispatch_batches: None
• split_batches: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• prompts: None
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	dim_384_cosine_ndcg@10	dim_256_cosine_ndcg@10	dim_128_cosine_ndcg@10	dim_64_cosine_ndcg@10
0.9474	9	-	0.1225	0.1221	0.1145	0.0915
1.0526	10	7.2521	-	-	-	-
2.0	19	-	0.1296	0.1261	0.1157	0.1089
2.1053	20	5.4977	-	-	-	-
2.9474	28	-	0.1294	0.1377	0.1262	0.1090
3.1579	30	4.3477	-	-	-	-
4.0	38	-	0.1330	0.1378	0.1260	0.1126
4.2105	40	3.3767	-	-	-	-
4.9474	47	-	0.1415	0.1388	0.1294	0.1221
5.2632	50	2.6443	-	-	-	-
6.0	57	-	0.1515	0.1395	0.1348	0.1218
6.3158	60	2.0824	-	-	-	-
6.9474	66	-	0.1480	0.1411	0.1335	0.1242
7.3684	70	1.6734	-	-	-	-
8.0	76	-	0.1491	0.1481	0.1428	0.1313
8.4211	80	1.3894	-	-	-	-
8.9474	85	-	0.1449	0.1497	0.1419	0.1341
9.4737	90	1.1443	-	-	-	-
10.0	95	-	0.1466	0.1494	0.1399	0.1396
10.5263	100	1.0121	-	-	-	-
10.9474	104	-	0.1458	0.1477	0.1415	0.1371
11.5789	110	0.8833	-	-	-	-
12.0	114	-	0.1479	0.1474	0.1445	0.1374
12.6316	120	0.8201	-	-	-	-
12.9474	123	-	0.1519	0.1486	0.1458	0.1360
13.6842	130	0.736	-	-	-	-
14.0	133	-	0.1505	0.1471	0.1484	0.1376
14.7368	140	0.6924	-	-	-	-
14.9474	142	-	0.1496	0.1486	0.1451	0.1396
15.7895	150	0.672	-	-	-	-
16.0	152	-	0.1492	0.1489	0.1464	0.1404
16.8421	160	0.6455	-	-	-	-
16.9474	161	-	0.1496	0.1493	0.1468	0.1389
17.8947	170	0.6538	-	-	-	-
18.0	171	-	0.1501	0.1470	0.1461	0.1393
18.9474	180	0.628	0.1501	0.1489	0.1465	0.1390

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 3.3.1
• Transformers: 4.41.2
• PyTorch: 2.1.2+cu121
• Accelerate: 0.34.2
• Datasets: 2.19.1
• Tokenizers: 0.19.1

Citation

BibTeX

Sentence Transformers

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}

MatryoshkaLoss

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}